Faiyaz Doctor currently works at the School of Computer Science and Electronic Engineering, University of Essex, United Kingdom
Skills and Expertise
Research Items (73)
Today, Intelligent Transportation Systems (ITS) have emerged to enhance the efficiency of existing transportation schemes by providing greater services and efficiencies while enhancing safety and security in the mobility of people and merchandise. One of the evolving research topics in ITS is Internet of Vehicles (IoV) which aims to form a large network of vehicles and other wireless elements and infrastructures. Autonomous vehicles are also an important aspect of ITS providing real-time sensing and detection of safety threats, optimizing route / task planning, power consumption, coordination and collaboration with other vehicles while performing smart self-diagnostics and maintenance under different operating environment. A natural evolution of IoVs is the Internet of Autonomous Vehicles (IAV) which extends the operational benefits of connected vehicles through incorporating intelligent real-time information processing, adaptation, self-learning and self-healing capabilities. In such system, Unmanned Areial Vehicles (UAV) autonomous underwater vehicle (AUV) could support surveillance and monitoring, search and rescue and anomaly and threat detection as well as accurate inspection, characterization tasks that are communicated to people or other terrestrial devices and robots.. However, there are several open research issues concerning the effective employment of multiple connected heterogeneous autonomous vehicles and devices for IoV applications such as: efficient and smart management of the autonomous vehicles and devices, reliable and secure communication between heterogeneous autonomous vehicles and devices and connectivity between autonomous vehicles. Due to the dynamic nature of the environment, operating conditions and mission objectives, balancing these issues becomes a challenging task. Hence, there is a need to develop novel techniques to manage and optimise real-time operation of these complex communication platforms in the face of dynamic and uncertain real-world deployment scenarios. This workshop aims at bringing together scholars from academia and industry to discuss and present the latest research and findings on all the aspects of IAVs.
To assist in the broadcasting of time-critical traffic information in an Internet of Vehicles (IoV) and vehicular sensor networks (VSN), fast network connectivity is needed. Accurate traffic information prediction can improve traffic congestion and operation efficiency, which helps to reduce commute times, noise and carbon emissions. In this study, we present a novel approach for predicting the traffic flow volume by using traffic data in self-organizing vehicular networks. The proposed method is based on using a probabilistic generative neural network techniques called deep belief network (DBN) that includes multiple layers of restricted Boltzmann machine (RBM) auto-encoders. Time series data generated from the roadside units (RSUs) for five highway links are used by a three layer DBN to extract and learn key input features for constructing a model to predict traffic flow. Back-propagation is utilized as a general learning algorithm for fine-tuning the weight parameters among the visible and hidden layers of RBMs. During the training process the firefly algorithm (FFA) is applied for optimizing the DBN topology and learning rate parameter. Monte Carlo simulations are used to assess the accuracy of the prediction model. The results show that the proposed model achieves superior performance accuracy for predicting traffic flow in comparison with other approaches applied in the literature. The proposed approach can help to solve the problem of traffic congestion, and provide guidance and advice for road users and traffic regulators.
Stress and sadness are some of the factors that affect the quality of life of people. Inadequate attention to health and safety regulations at work causes about 2 million work-related deaths annually. In this paper, we propose a novel architecture able to monitor the emotional state of the user and allow personalized feedback through a smart music player application. There are research initiatives aiming to predict, detect and intervene when dangerous conditions are found, although there is little research focused on obtaining and processing emotional information obtained from social networks. The architecture proposed can detect and handle the emotional state of the user and offer personalized music therapy to improve their mood. The emotional state is detected analyzing the tweets from the user, and the IBM Tone Analyzer is used as a tool for processing the information. Preliminary experiments were performed with encouraging results.
Question - Salable Implementations of Parameterised Markov Models?
Thanks for this, will have a look, have you used this before?
In this paper, we present a novel system for cognitive stimulation therapy to progressively assess cognitive impairment and emotional well-being of dementia patients in social care settings. The system assesses patients interactions and computes performance scores for different areas of cognitive stimulation. Patient interactions are initially classified into predefined performance categories through clustering of a sampled population. New personalized stimulation plans tailored to match the patient's changing level of impairment are generated automatically through a set of fuzzy rule based systems using quantitative attributes and the overall scores of patients interactions. Therapists can redefine, evaluate and adjust the rules governing difficulty and activity levels for different stimulation areas to fine tune generated activity plans. The system can also be combined with an Internet of Things (IoT) enabled patient dialogue system for determining the affective state of participants during therapy sessions that could be used as a pervasive condition monitoring platform. Experiments consisting of therapy sessions of patients interacting with the system were performed in which the activity plans were automatically generated. Initial results showed that the system outputs were in agreement with the therapists own assessment in most of the stimulation areas. Simulation experiments were also conducted to analyse the system performance over multiple sessions. The results suggest that the system is able to adapt therapy plans overtime in response to changing levels of impairment/performance while supporting therapists to tune and evaluate therapy plans more effectively.
With rapid development of computer vision and artificial intelligence, cities are becoming more and more intelligent. Recently, since intelligent surveillance was applied in all kind of smart city services, object tracking attracted more attention. However, two serious problems blocked development of visual tracking in real applications. The first problem is its lower performance under intense illumination variation while the second issue is its slow speed. This paper addressed these two problems by proposing a correlation filter based tracker. Fog computing platform was deployed to accelerate the proposed tracking approach. The tracker was constructed by multiple positions' detections and alternate templates (MPAT). The detection position was repositioned according to the estimated speed of target by optical flow method, and the alternate template was stored with a template update mechanism, which were all computed at the edge. Experimental results on large-scale public benchmark datasets showed the effectiveness of the proposed method in comparison with state-of-the-art methods.
- May 2018
During surgical procedures, bispectral index (BIS) is a well-known measure used to determine the patient’s depth of anesthesia (DOA). However, BIS readings can be subject to interference from many factors during surgery, and other parameters such as blood pressure (BP) and heart rate (HR) can provide more stable indicators. However, anesthesiologist still consider BIS as a primary measure to determine if the patient is correctly anaesthetized while relaying on the other physiological parameters to monitor and ensure the patient’s status is maintained. The automatic control of administering anesthesia using intelligent control systems has been the subject of recent research in order to alleviate the burden on the anesthetist to manually adjust drug dosage in response physiological changes for sustaining DOA. A system proposed for the automatic control of anesthesia based on type-2 Self Organizing Fuzzy Logic Controllers (T2-SOFLCs) has been shown to be effective in the control of DOA under simulated scenarios while contending with uncertainties due to signal noise and dynamic changes in pharmacodynamics (PD) and pharmacokinetic (PK) effects of the drug on the body. This study considers both BIS and BP as part of an adaptive automatic control scheme, which can adjust to the monitoring of either parameter in response to changes in the availability and reliability of BIS signals during surgery. The simulation of different control schemes using BIS data obtained during real surgical procedures to emulate noise and interference factors have been conducted. The use of either or both combined parameters for controlling the delivery Propofol to maintain safe target set points for DOA are evaluated. The results show that combing BIS and BP based on the proposed adaptive control scheme can ensure the target set points and the correct amount of drug in the body is maintained even with the intermittent loss of BIS signal that could otherwise disrupt an automated control system.
- Apr 2018
Big Data has significant impact in developing functional smart cities and supporting modern societies. In this paper, we investigate the importance of Big Data in modern life and economy, and discuss challenges arising from Big Data utilization. Different computational intelligence techniques have been considered as tools for Big Data analytics. We also explore the powerful combination of Big Data and Computational Intelligence (CI) and identify a number of areas, where novel applications in real world smart city problems can be developed by utilizing these powerful tools and techniques. We present a case study for intelligent transportation in the context of a smart city, and a novel data modelling methodology based on a biologically inspired universal generative modelling approach called Hierarchical Spatial-Temporal State Machine (HSTSM). We further discuss various implications of policy, protection, valuation and commercialization related to Big Data, its applications and deployment.
Project - Intelligent Urban Flood Prediction and Monitoring for Malaysia
The project is currently seeking to appoint two data scientists with technical and business skills to help to realise the outcomes and deliverables of the project. Further information about the appointments can be found on the attached documents.
Question - Where can I obtain a SAS Macro Program for Modelling Multi-State Markov Processes?
Thanks but I have already seem the paper, I cannot get the source code as the website mentioned in the paper is no longer working and the authors and not reachable.
- Jan 2018
- Advances in Hybridization of Intelligent Methods
Smart city combines connected services from different disciplines offering a promise of increased efficiency in transport and mobility in urban environment. This has been enabled through many important advancements in fields like machine learning, big data analytics, hardware manufacturing and communication technology. Especially important in this context is big data which is fueling the digital revolution in an increasingly knowledge driven society by offering intelligence solutions for the smart city. In this paper, we discuss the importance of big data analytics and computational intelligence techniques for the problem of taxi traffic modelling, visualisation and prediction. This work provides a comprehensive survey of computational intelligence techniques appropriate for the effective processing and analysis of big data. A brief description of many smart city projects, initiatives and challenges in the UK is also presented. We present a hybrid data modelling approach used for the modelling and prediction of taxi usage. The approach introduces a novel biologically inspired universal generative modelling technique called Hierarchical Spatial-Temporal State Machine (HSTSM). The HSTSM modelling approach incorporates many soft computing techniques including: deep belief networks, auto-encoders, agglomerative hierarchical clustering and temporal sequence processing. A case study for the modelling and prediction of traffic based on taxi movements is described, where HSTSM is used to address the computational challenges arising from analysing and processing large volumes of varied data.
- Dec 2017
Fast evolution of the Internet, mobile technologies and energy efficient communication protocols has given a new momentum to e-businesses and world become a global village. Due to increase in the usage of internet and number of mobile users, many companies use these channels to make their products and brands visible to their customers all over the world. Although, some progress has been made towards this direction but further exploration is required, particularly it is still a challenge to enhance in-flight passengers’ shopping experience through efficient and reliable communication protocols. In this paper, we proposed a framework for omni-channel which is based on cognitive radio and machine learning. The proposed cognitive radio communication protocols provide seamless connectivity to in-flight passengers through energy efficient mode like machine learning (ML). Here, machine learning helps to develop user profile, based on relevance feedback that address the problem of catalogue and information overload. In this paper, we also discuss various challenges and opportunities associated with the proposed omni-channel business model. Moreover, the role and impact of emerging technologies such as cognitive radio and 5G in realizing omni-channel businesses is discussed in this paper. Our results explain the seamless communication between aircraft users and merchandise, through reliable and efficient connectivity when the aircraft passes over different geographic areas i.e. urban/rural land or sea at different altitudes and geographic locations. Here, backup data channel is introduced which further enhance the reliability of connection especially when primary users turns ON during the communication. Furthermore, the proposed model helps to reduce communication time and consume less energy to transmit with high throughput as compared to the benchmark cognitive radio protocols.
- Nov 2017
Big data is fuelling the digital revolution in an increasingly knowledge driven and connected society by offering big data analytics and computational intelligence based solutions to reduce the complexity and cognitive burden on accessing and processing large volumes of data. In this paper, we discuss the importance of big data analytics and computational intelligence techniques applied to data produced from the myriad of pervasively connected machines and personalized devices offering embedded and distributed information processing capabilities. We provide a comprehensive survey of computational intelligence techniques appropriate for the effective processing and analysis of big data. We discuss a number of exemplar application areas that generate big data and can hence benefit from its effective processing. State of the art research and novel applications in health-care, intelligent transportation and social network sentiment analysis, are presented and discussed in the context of Big data, Cyber-Physical Systems (CPS), and Computational Intelligence (CI). We present a data modelling methodology, which introduces a novel biologically inspired universal generative modelling approach called Hierarchical Spatial-Temporal State Machine (HSTSM). The HSTSM modelling approach incorporates a number of soft computing techniques such as: deep belief networks, auto-encoders, agglomerative hierarchical clustering and temporal sequence processing, in order to address the computational challenges arising from analysing and processing large volumes of diverse data to provide an effective big data analytics tool for diverse application areas. A conceptual cyber-physical architecture, which can accommodate and benefit from the proposed methodology, is further presented.
- Jul 2017
- IEEE International conference on Fuzzy systems
Fuzzy Cognitive Maps (FCMs) are a soft computing technique characterized by robust properties that make them an effective technique for medical decision support systems. Making decisions within a medical domain is difficult due to the existence of high levels of uncertainty. The sources of this uncertainty can be due to the variation of physicians' opinions and experiences. The structure of existing FCMs is based on type -1 fuzzy sets in order to represent the causal relations among concepts of the modeled system. Therefore, the ability of the FCM to handle high levels of uncertainties and deliver accurate results can be hindered. In this paper, we propose using the Interval Agreement Approach to model the weights of links in FCMs to capture high level uncertainties in the presence of imprecise data acquired from different medical experts to enhance its decision modelling and reasoning capability. The proposed model is used in identifying if a child is diagnosed with an Autism Spectrum Disorder (ASD) where the Modified Checklist for Autism in Toddlers is used as a standard tool to derive the inputs for the FCMs. Initial results demonstrate that the proposed method outperforms conventional FCMs in classifying ASD based on a dataset of diagnosed cases.
Project - Data Driven Computational Models for Prediction and Simulation of Complex Dynamic Labour Market Systems
The post-doctoral position is now advertised online at:
Please kindly circulate within your networks, application deadline is 13th April 2017.
- Feb 2017
This paper presents a novel emotion modeling methodology for incorporating human emotion into intelligent computer systems. The proposed approach includes a method to elicit emotion information from users, a new representation of emotion (AV-AT model) that is modelled using a genetically optimized adaptive fuzzy logic technique, and a framework for predicting and tracking user's affective trajectory over time. The fuzzy technique is evaluated in terms of its ability to model affective states in comparison to other existing machine learning approaches. The performance of the proposed affect modeling methodology is tested through the deployment of a personalised learning system, and series of offline and online experiments. A hybrid cloud intelligence infrastructure is used to conduct large-scale experiments to analyze user sentiments and associated emotions, using data from a million Facebook users. A performance analysis of the infrastructure on processing, analyzing, and data storage has been carried out, illustrating its viability for large-scale data processing tasks. A comparison of the proposed emotion categorizing approach with Facebook's sentiment analysis API demonstrates that our approach can achieve comparable performance. Finally, discussions on research contributions to cloud intelligence using sentiment analysis, emotion modeling, big data, and comparisons with other approaches are presented in detail.
- Jan 2017
With the exponential growth of information available on the Internet and various organisational intranets there is a need for profile based information seeking and retrieval (IS&R) systems. These systems should be able to support users with their context-aware information needs. This paper presents a new approach for enterprise IS&R systems using fuzzy logic to develop task, user and document profiles to model user information seeking behaviour. Relevance feedback was captured from real users engaged in IS&R tasks. The feedback was used to develop a linear regression model for predicting document relevancy based on implicit relevance indicators. Fuzzy relevance profiles were created using Term Frequency and Inverse Document Frequency (TF-IDF) analysis for the successful user queries. Fuzzy rule based summarisation was used to integrate the three profiles into a unified index reflecting the semantic weight of the query terms related to the task, user and document. The unified index was used to select the most relevant documents and experts related to the query topic. The overall performance of the system was evaluated based on standard precision and recall metrics which show significant improvements in retrieving relevant documents in response to user queries.
- Dec 2016
- 2016 IEEE Symposium Series on Computational Intelligence (SSCI)
There is a prominent connection between human health and human emotion. This connection has encouraged researchers to produce numerous applications in order to facilitate patients and therapists. In this paper, through a review study we highlight how the development of intelligent emotion aware e-health systems can facilitate patient's satisfaction, emotion wellbeing, and physical health, and improve the quality of service offered by health care related businesses. Moreover, we discuss the challenges and difficulties concerning emotion recognition and modelling systems responsible for representing the patient's affective state in real life health care environments. In our research, we aim to address these challenges by proposing a novel framework for developing emotion aware health care support systems. The suggested methodology enables a holistic and reflective representation of the patient's affective state, and incorporates a number of design choices that are suitable for emotion modelling and recognition in the context of a real life health care environment. This methodology leads to the development of a unique emotion aware health care support system, which utilizes Fuzzy Logic to recognize the patient's affective state based on basic cognitive/affective cues, such as the patient's predictions and evaluations of a treatment. The system based on the calculated emotion recognition results, delivers tailored feedback to influence the patient towards a desired and beneficial affective state. As demonstrated in this paper, the proposed emotion-modelling methodology could be very useful when applied in specific real life contexts to develop novel health care systems that are able to accurately monitor and predict their user's emotions.
Rigorous analysis of user interest in web documents is essential for the development of recommender systems. This paper investigates the relationship between the implicit parameters and user explicit rating during their search and reading tasks. The objective of this paper is therefore three-fold: firstly, the paper identifies the implicit parameters which are statistically correlated with the user explicit rating through user study 1. These parameters are used to develop a predictive model which can be used to represent users’ perceived relevance of documents. Secondly, it investigates the reliability and validity of the predictive model by comparing it with eye gaze during a reading task through user study 2. Our findings suggest that there is no significant difference between the predictive model based on implicit indicators and eye gaze within the context examined. Thirdly, we measured the consistency of user explicit rating in both studies and found significant consistency in user explicit rating of document relevance and interest level which further validates the predictive model. We envisage that the results presented in this paper can help to develop recommender and personalised systems for recommending documents to users based on their previous interaction with the system.
- Mar 2016
- The first International Conference on Internet of Things and Big Data, special session, Recent Advancement in Internet of Things, Big Data and Security (RAIBS)
In this paper we present a novel affective modelling approach to be utilised by Affective Computing systems. This approach is a combination of the well known Arousal Valence model of emotion and the newly introduced Affective Trajectories Hypothesis. An adaptive data driven fuzzy method is proposed in order to extract personalized emotion models, and successfully visualise the associations of these models’ basic elements, to different emotional labels, using easily interpretable fuzzy rules. Namely we explore how the combinations of arousal, valence, prediction of the future, and the experienced outcome after this prediction, enable us to differentiate between different emotional labels. We use the results obtained from a user study consisting of an online survey, to demonstrate the potential applicability of this affective modelling approach, and test the effectiveness and stability of its adaptive element, which accounts for individual differences between the users. We also propose a basic architecture in order for this approach to be used effectively by AC systems, and finally we present an implementation of a personalised learning system which utilises the suggested framework. This implementation is tested through a pilot experimental session consisting of a tutorial on fuzzy logic which was conducted under an activity-led and problem based learning context.
- Dec 2015
In this paper, we present a data analytics and visualization framework for health-shocks prediction based on large-scale health informatics dataset. The framework is developed using cloud computing services based on Amazon web services (AWS) integrated with geographical information systems (GIS) to facilitate big data capture, storage, index and visualization of data through smart devices for different stakeholders. In order to develop a predictive model for health-shocks, we have collected a unique data from 1000 households, in rural and remotely accessible regions of Pakistan, focusing on factors like health, social, economic, environment and accessibility to healthcare facilities. We have used the collected data to generate a predictive model of health-shock using a fuzzy rule summarization technique, which can provide stakeholders with interpretable linguistic rules to explain the causal factors affecting health-shocks. The evaluation of the proposed system in terms of the interpret-ability and accuracy of the generated data models for classifying health-shock shows promising results. The prediction accuracy of the fuzzy model based on a k-fold cross-validation of the data samples shows above 89% performance in predicting health-shocks based on the given factors.
- Nov 2015
- 2015 Conference on Technologies and Applications of Artificial Intelligence (TAAI)
In this paper, a novel genetic type-2 self-organising fuzzy logic controller (SOFLC) is proposed for anaesthesia control. The genetic type-2 SOFLC has a hierarchical structure consisting of three layers: a basic type-2 fuzzy logic controller, a self-organising mechanism for online adaption, and a genetic algorithm for offline optimisation. The genetic type-2 SOFLC is tested under different levels of environmental noise and compared with basic type-2 SOFLC that does not optimised. The results show that the proposed genetic type-2 SOFLC can perform better than the type-2 SOFLC in the presence of noise in terms of steady state error.
The importance of information retrieval systems is unquestionable in the modern society and both individuals as well as enterprises recognise the benefits of being able to find information effectively. Current code-focused information retrieval systems such as Google Code Search, Codeplex or Koders produce results based on specific keywords. However, these systems do not take into account developers’ context such as development language, technology framework, goal of the project, project complexity and developer’s domain expertise. They also impose additional cognitive burden on users in switching between different interfaces and clicking through to find the relevant code. Hence, they are not used by software developers. In this paper, we discuss how software engineers interact with information and general-purpose information retrieval systems (e.g. Google, Yahoo!) and investigate to what extent domain-specific search and recommendation utilities can be developed in order to support their work-related activities. In order to investigate this, we conducted a user study and found that software engineers followed many identifiable and repeatable work tasks and behaviours. These behaviours can be used to develop implicit relevance feedback-based systems based on the observed retention actions. Moreover, we discuss the implications for the development of task-specific search and collaborative recommendation utilities embedded with the Google standard search engine and Microsoft IntelliSense for retrieval and re-engineering of code. Based on implicit relevance feedback, we have implemented a prototype of the proposed collaborative recommendation system, which was evaluated in a controlled environment simulating the real-world situation of professional software engineers. The evaluation has achieved promising initial results on the precision and recall performance of the system.
In this paper, novel interval and general type-2 self-organizing fuzzy logic controllers (SOFLCs) are proposed for the automatic control of anesthesia during surgical procedures. The type-2 SOFLC is a hierarchical adaptive fuzzy controller able to generate and modify its rule-base in response to the controller's performance. The type-2 SOFLC uses type-2 fuzzy sets derived from real surgical data capturing patient variability in monitored physiological parameters during anesthetic sedation, which are used to define the footprint of uncertainty (FOU) of the type-2 fuzzy sets. Experimental simulations were carried out to evaluate the performance of the type-2 SOFLCs in their ability to control anesthetic delivery rates for maintaining desired physiological set points for anesthesia (muscle relaxation and blood pressure) under signal and patient noise. Results show that the type-2 SOFLCs can perform well and outperform previous type-1 SOFLC and comparative approaches for anesthesia control producing lower performance errors while using better defined rules in regulating anesthesia set points while handling the control uncertainties. The results are further supported by statistical analysis which also show that zSlices general type-2 SOFLCs are able to outperform interval type-2 SOFLC in terms of their steady state performance.
The importance of finding relevant information for business and decision making is imperative for both individuals as well as enterprises. In this paper, we present an approach for the development of a fuzzy information retrieval (IR) system. The approach provides a new mechanism for constructing and integrating three relevancy profiles: a task profile, a user profile and document profile, into a unified index through the use of relevance feedback and fuzzy rule based summarisation. Experiments were performed from which relevance feedback and user queries were captured from 35 users on 20 predefined simulated enterprise search tasks. The captured data set was used to develop the three types of profiles and train the fuzzy system. The system shows 86% performance accuracy in correctly classifying document relevance. The overall performance of the system was evaluated based on standard precision and recall which shows significant improvements in retrieving relevant documents based on user queries.
- Mar 2015
Depth of anaesthesia (DoA) is an important measure for assessing the degree to which the central nervous system of a patient is depressed by a general anaesthetic agent, depending on the potency and concentration with which anaesthesia is administered during surgery. We can monitor the DoA by observing the patient's electroencephalography (EEG) signals during the surgical procedure. Typically high frequency EEG signals indicates the patient is conscious, while low frequency signals mean the patient is in a general anaesthetic state. If the anaesthetist is able to observe the instantaneous frequency changes of the patient's EEG signals during surgery this can help to better regulate and monitor DoA, reducing surgical and post-operative risks. This paper describes an approach towards the development of a 3D real-time visualization application which can show the instantaneous frequency and instantaneous amplitude of EEG simultaneously by using empirical mode decomposition (EMD) and the Hilbert-Huang transform (HHT). HHT uses the EMD method to decompose a signal into so-called intrinsic mode functions (IMFs). The Hilbert spectral analysis method is then used to obtain instantaneous frequency data. The HHT provides a new method of analyzing non-stationary and nonlinear time series data. We investigate this approach by analyzing EEG data collected from patients undergoing surgical procedures. The results show that the EEG differences between three distinct surgical stages computed by using sample entropy (SampEn) are consistent with the expected differences between these stages based on the bispectral index (BIS), which has been shown to be a quantifiable measure of the effect of anaesthetics on the central nervous system. Also, the proposed filtering approach is more effective compared to the standard filtering method in filtering out signal noise resulting in more consistent results than those provided by the BIS. The proposed approach is therefore able to distinguish between key operational stages related to DoA, which is consistent with the clinical observations. SampEn can also be viewed as a useful index for evaluating and monitoring the DoA of a patient when used in combination with this approach.
- Feb 2015
This paper presents a novel approach to the detection and recognition of faulty audio signalling devices as part of an automated industrial manufacturing quality assurance process. The proposed system outperforms other well-established automated systems based on mel-frequency cepstrum coefficients (MFCC) and multi-layer perceptron (MLP). It uses both unlabelled sound data and labelled historical data acquired from human experts in detecting faulty signalling devices. The unlabelled data is used to train a deep neural network generative model to create multiple levels of hierarchical feature extractors which are used to train an MLP classifier, with the intent to model the human reasoning and judging processes in respect to sound classification. This paper presents the results of real world experiments based on data pertaining to the audio signalling quality assurance process for car instrument cluster manufacturing. These results show that the proposed system is able to successfully classify speakers into two groups: “Good” and “No good” depending on the part quality. The proposed system proves to be capable enough to eliminate the need for a manual inspection within the manufacturing process and is shown to be able to diagnose a fault with a high degree of accuracy. This work can be extended to other areas of automotive inspection where there is a need for a robust solution to sound detection and where an output signal is represented by a complex and changing frequency spectrum even with significant environmental noise.
We compare type-1 and type-2 self-organizing fuzzy logic controller (SOFLC) using expert initialized and pretrained extracted rule-bases applied to automatic control of anaesthesia during surgery. We perform experimental simulations using a nonfixed patient model and signal noise to account for environmental and patient drug interaction uncertainties. The simulations evaluate the performance of the SOFLCs in their ability to control anesthetic delivery rates for maintaining desired physiological set points for muscle relaxation and blood pressure during a multistage surgical procedure. The performances of the SOFLCs are evaluated by measuring the steady state errors and control stabilities which indicate the accuracy and precision of control task. Two sets of comparisons based on using expert derived and extracted rule-bases are implemented as Wilcoxon signed-rank tests. Results indicate that type-2 SOFLCs outperform type-1 SOFLC while handling the various sources of uncertainties. SOFLCs using the extracted rules are also shown to outperform those using expert derived rules in terms of improved control stability.
In this paper, we discuss the development of an ambient intelligent-based system for the monitoring of dementia patients living in their own homes. Within this system groups of unobtrusive wireless sensor devices can be deployed at specific locations within a patient’s home and accessed via standardized interfaces provided through an open middleware platform. For each sensor group intelligent agents are used to learn fuzzy rules, which model the patient’s habitual behaviours in the environment. An online rule adaptation technique is applied to facilitate short-term tuning of the learnt behaviours, and long-term tracking of behaviour changes which could be due to the effects of cognitive decline caused from dementia. The proposed system reports macro level behaviour changes and micro level perception drift to care providers to enable them to make better-informed assessments of the patient’s cognitive abilities and changing care needs. We demonstrate experiments in a real pervasive computing environment, in which our intelligent agent approach can learn to model the user’s behaviours and allow online adaptation of its model to better approximate the learnt behaviours and identify long-term macro-level behaviour changes, which could be attributed to cognitive decline. We also show an example of how the user’s perceptions for thermal comfort may be captured and visualised to provide a means by which micro-level perception changes can be monitored.
- Oct 2014
In this paper we propose a SoPC-based multiprocessor embedded system for controlling ambiental parameters in an Intelligent Inhabited Environment. The intelligent features are achieved by means of a Neuro-Fuzzy system which has the ability to learn from samples, reason and adapt itself to changes in the environment or in user preferences. In particular, a modified version of the well known ANFIS (Adaptive Neuro-Fuzzy Inference System) scheme is used, which allows the development of very efficient implementations. The architecture proposed here is based on two soft-core microprocessors: one microprocessor is dedicated to the learning and adaptive procedures, whereas the other is dedicated to the on-line response. This second microprocessor is endowed with 4 efficient ad-hoc hardware modules intended to accelerate the neuro-fuzzy algorithms. The implementation has been carried out on a Xilinx Virtex-5 FPGA and obtained results show that a very high performance system is achieved.
- Sep 2014
- 10th IFIP WG 12.5 International Conference, AIAI 2014
Humans vary in their learning behaviour. It is difficult to predict the actual needs of learners through their search activity. It is also difficult to predict accurately the level of satisfaction after the learner finds a perceived relevant document. This research is a preliminary study to examine the predictive strength of some implicit indicators on web documents. An automated study was carried out and 13 participants were given 15 short documents to read and rate according to their perception of relevance to a given topic area. An investigation was carried out to examine if there exists a correlation between user generated implicit indicators and the explicit ratings. The findings show that there is a positive correlation between the dwell time and user explicit ratings. Although there was no significant correlation between mouse movement/distance and user explicit rating, there was a relationship between the homogeneous clusters of the implicit indicators and the user ratings.
- Jul 2014
- 2014 International Joint Conference on Neural Networks (IJCNN)
The availability of advanced driver assistance systems (ADAS), for safety and well-being, is becoming increasingly important for avoiding traffic accidents caused by fatigue, stress, or distractions. For this reason, automatic identification of a driver from among a group of various drivers (i.e. real-time driver identification) is a key factor in the development of ADAS, mainly when the driver's comfort and security is also to be taken into account. The main focus of this work is the development of embedded electronic systems for in-vehicle deployment of driver identification models. We developed a hybrid model based on artificial neural networks (ANN), and cepstral feature extraction techniques, able to recognize the driving style of different drivers. Results obtained show that the system is able to perform real-time driver identification using non-intrusive driving behavior signals such as brake pedal signals and gas pedal signals. The identification of a driver from within groups with a reduced number of drivers yields promising identification rates (e.g. 3-driver group yield 84.6 %). However, real-time development of ADAS requires very fast electronic systems. To this end, an FPGA-based hardware coprocessor for acceleration of the neural classifier has been developed. The coprocessor core is able to compute the whole ANN in less than 4 μs.
- Jun 2014
- 2014 International Conference on Intelligent Environments (IE)
In recent years, the problem of cyclic instability has been investigated mainly using two approaches: analysing the topological properties of the system (finding loops or feedback) and bio-inspired optimization. One of the main disadvantages of analysing the topology of the system (i.e. The connectivity of the agents involved in the environment) is the computational cost (that could be increased if the environment includes nomadic agents). Optimization-based approaches have been proven to work very well, even in the case of nomadic agents. However, the optimisation approach has been deployed mainly using computer simulations. With the breakthrough of integrated circuits, allowing a wide variety of low cost microcontrollers, the possibility of implementing intelligent algorithms (such as fuzzy logic, neural networks, etc.) on embedded agents is a reality. In this paper, we present a preliminary analysis toward the implementation of bio-inspired optimisation algorithms on embedded systems. Our long-term goal is to be able to prevent cyclic instability in real and complex rule based multi-agent environments using optimisation algorithms on embedded system.
Monitoring students' activity and performance is vital to enable educators to provide effective teaching and learning to engage students with the subject and improve their understanding of the material. We describe the use of a fuzzy linguistic summarisation (LS) technique for extracting linguistically interpretable rules from student data describing prominent relationships between activity/engagement characteristics and achieved performance. We propose an intelligent framework for monitoring individual or group performance during activity and problem-based learning tasks. The proposed system is developed as a set of services to cater for data heterogeneity and deployable on a cloud computing platform. We present a case study and experiments in which we apply the fuzzy LS technique for analysing the effectiveness of using a group performance model (GPM) to deploy activity led learning (ALL) in a master-level module. Results show that the fuzzy rules can identify useful relationships between student engagement and performance.
- Dec 2013
- Proceedings of the 2013 Conference on Technologies and Applications of Artificial Intelligence
In this paper, three different self-organizing fuzzy logic controllers (SOFLC), type-1, interval type-2 and zSlice general type-2 are used in the simulated control of anesthesia. Experimental results show that both the type-2 SOFLCs have better performance in comparison to the type-1 system. Based on the performance results, we then select the interval type-2 SOFLC to analyze rule usage during real-time regulation and control of anesthesia. Based on this we extracted a new reduced rule base which can provide a better basis to further research in understanding the SOFLC dynamic control behavior.
- Oct 2013
- Proceedings of the 2013 IEEE International Conference on Systems, Man, and Cybernetics
This paper presents a novel system for automatic detection and recognition of faulty audio signaling devices as part of an automated industrial manufacturing process. The system uses historical data labeled by human experts in detecting faulty signaling devices to train an artificial neural network based classifier for modeling their decision making process. The neural network is implemented on a real time embedded micro controller which can be more efficiently incorporated into an automated production line eliminating the need for a manual inspection within the manufacturing process. We present real world experiments based on data pertaining to the production and manufacture of audio signaling components used in car instrument clusters. Our results show that the proposed expert system is able to successfully classify faulty audio signaling devices to a high degree of accuracy. The results can be generalized to other signaling devices where an output signal is represented by a complex and changing frequency spectrum even with significant environmental noise.
According to a first aspect of the invention there is provided a method of decision-making comprising: a data input step to input data from a plurality of first data sources into a first data bank, analysing said input data by means of a first adaptive artificial neural network (ANN), the neural network including a plurality of layers having at least an input layer, one or more hidden layers and an output layer, each layer comprising a plurality of interconnected neurons, the number of hidden neurons utilized being adaptive, the ANN determining the most important input data and defining therefrom a second ANN, deriving from the second ANN a plurality of Type-1 fuzzy sets for each first data source representing the data source, combining the Type-1 fuzzy sets to create Footprint of Uncertainty (FOU) for type-2 fuzzy sets, modelling the group decision of the combined first data sources; inputting data from a second data source, and assigning an aggregate score thereto, comparing the assigned aggregate score with a fuzzy set representing the group decision, and producing a decision therefrom. A method employing a developed ANN as defined in Claim 1 and extracting data from said ANN, the data used to learn the parameters of a normal Fuzzy Logic System (FLS).
The powerful synergy of neural networks and reconfigurable hardware provides a solid foundation for the development of high performance embedded systems able to efficiently adapt to changing requirements. Adaptation at different levels — ranging from the physical level to the system level-can be combined to develop efficient solutions by means of FPGA technology. In this work, a multilevel adaptation scheme for the development of intelligent agents is proposed. Software learning algorithms are applied to adapt the agent behavior (i.e. neural network parameters) at the system level, while dynamic partial reconfiguration (DPR) is used to modify the agent at the physical and architectural level (i.e. neural network topology). Firstly, a multilevel adaptive intelligent agent is able to manage its resources efficiently in order to meet time-varying demands such as speed performance and power consumption. Secondly, from the behavioral viewpoint, multilevel adaptation provides the intelligent agent with high plasticity and flexibility. An FPGA-based intelligent agent has been successfully deployed for a real-time control problem in an inhabited intelligent environment. Results obtained show that the agent is able to adapt itself to changes in the environment in a lifelong mode.
- Jun 2013
- Computer Supported Cooperative Work in Design (CSCWD), 2013 IEEE 17th International Conference on
- International Conference on Computer Supported Cooperative Work in Design
In order to enhance students' problem-solving competences it is necessary to develop their technical skill as well as their soft skills such as business, communication and team working. In this paper, we present an Activity-Led Learning (ALL) approach and analyze its impact on students' engagement (time-on-task), satisfaction and group performance. We propose a Group Performance Model (GPM) to deploy ALL effectively in the master-level, Network Planning and Management, module. The model provides a structure within which students are introduced to the ALL pedagogical methodology. The model systematically helps to facilitate group formation and allows group integration and cooperation by developing 'common ground' amongst group members. In order to evaluate the usage of GPM in ALL, we conducted group performance analysis using a fuzzy rule-based classification model. The results of the analysis showed that the application of GPM resulted in a reduction in overall time spent on tasks, while achieving better grades. This indicates that GPM can help groups to develop common ground, coordinate their activities and overcome inter-personal issues to achieve better overall performance in shorter times as opposed to groups in which GPM has not been applied. Students' direct feedback on module also shows the effectiveness of ALL on students' performance in general.
As the proportion of older adults grows, the number of special care provisions to help individuals with declining cognitive abilities needs to also increase. Information Communication Technology (ICT) is beginning to play an increasing role in facilitating the work of specialists to support and monitor individuals with cognitive impairment within their everyday environments. In addition, advances in artificial intelligence and the development of new algorithmic approaches can be used to approximate the computational processes of human behaviour in different circumstances. In this paper, we report on the development of a software system using game based therapies for older adults in Mexico suffering from cognitive impairment, where this system has been deployed in a unique day therapy centre. We further propose an evaluation module based on using AI approaches and affective sensing to monitor and detect significant changes in performance cognation that might indicate a possible cognitive decline.
- Jan 2013
- Mexican International Conference on Computer Science (ENC)
- Mexican International Conference on Computer Science
As the proportion of older adults grows, the number of special care provisions to help individuals with declining cognitive abilities needs also to increase. Information Communication Technology (ICT) is beginning to play an increasing role in facilitating the work and research of specialists to support and monitor individuals with cognitive impairment within their everyday environments. In addition, advances in artificial intelligence and the development of new algorithmic approaches can be used to approximate the computational processes of human behaviour in different circumstances. In this paper, we report on the development of a software system using game based therapies for older adults in Mexico suffering from cognitive impairment, where this system has been deployed in a unique day therapy centre. We further propose an evaluation module based on using an Artificial Neural Network (ANN) approach to monitor the user performance, a Fuzzy Logic based module to detect significant changes in performance that might indicate a possible cogni-tive decline, and a bio-signals sensor in order to gather information about the emotional state of the patient during the interaction .
- Sep 2012
- International Conference on Engineering Applications of Neural Networks
This paper investigates the development of a system for monitoring of dementia suffers living in their own homes. The system uses unobtrusive pervasive sensor and actuator devices that can be deployed within a patient's home grouped and accessed via standardized platforms. For each sensor group our system uses unsupervised neural networks to identify the patient's habitual behaviours based on their activities in the environment. Rule-based summarisation is used to provide descriptive rules representing the intra and inter activity variations within the discovered behaviours. We propose a model comparison mechanism to facilitate tracking of behaviour changes, which could be due to the effects of cognitive decline. We demonstrate using user data acquired from a real pervasive computing environment, how our system is able to identify the user's prominent behaviours enabling assessment and future tracking.
This paper presents the development of an embedded intelligent agent able to perform real-time control of ambientintelligence environments. The system has been implemented as a system-on-programmable chip (SoPC) on a field programmable gate array (FPGA). The scheme used for realizing the intelligent agent is an adaptive neuro-fuzzy system (NFS) enhanced with a principal component analysis (PCA) pre-processor. The PCA pre-processing stage allows a reduction of the input dimensions (features) with no meaningful loss of modeling capability. As a consequence, the computational complexity of the system is significantly reduced, allowing its implementation on a single electronic device. The NFS-PCA agent has been tested with data obtained in a real ubiquitous computing environment test bed.Results obtained show that the agent is able to perform real-time control of the environment in a proactive and non-intrusive way, and also to adapt to changes of user’s preferences in a life-long mode.
Monitoring students' activity and performance is vital to enable educators to provide effective teaching and learning in order to better engage students with the subject and improve their understanding of the material being taught. We describe the use of a fuzzy Linguistic Summarisation (LS) technique for extracting linguistically interpretable scaled fuzzy weighted rules from student data describing prominent relationships between activity / engagement characteristics and achieved performance. We propose an intelligent framework for monitoring individual or group performance during activity and problem based learning tasks. The system can be used to more effectively evaluate new teaching approaches and methodologies, identify weaknesses and provide more personalised feedback on learner's progress. We present a case study and initial experiments in which we apply the fuzzy LS technique for analysing the effectiveness of using a Group Performance Model (GPM) to deploy Activity Led Learning (ALL) in a Master-level module. Results show that the fuzzy weighted rules can identify useful relationships between student engagement and performance providing a mechanism allowing educators to transparently evaluate teaching and factors effecting student performance, which can be incorporated as part of an automated intelligent analysis and feedback system.
- May 2012
Even though there exists a number of search solutions targetted at software engineers the literature suggests that they are not widely used by the people engaged in code delivery . Moreover, current code focused information retrieval systems such as Google Code Search (discontinued), Codeplex or Koders produce results based on specific keywords and therefore they do not take into account user context such as location, browsing history, previous interaction patterns and domain expertise. In this paper we discuss the development of task-specific information retrieval systems for software engineers. We discuss how software engineers interact with information and information retrieval systems and investigate to what extent a domain-specific search and recommendation system can be developed in order to support their work related activities. We have conducted a user study: a questionnaire and an automated observation of user interactions with the browser and software development environment. We discuss factors that can be used as implicit feedback indicators for further collaborative filtering and discuss how these parameters can be analysed using Computational Intelligence based techniques.
Variations in treatment decisions made by individual physicians can lead to practice variations in which different doctors may treat patients with the same type of medical conditions differently. There is a need to develop novel decision support systems that can interpret the decision-making behaviour of clinicians and identify the factors influencing particular treatment decisions. Such a tool will ensure that patients with the same medical conditions consistently receive the same kind of treatment and care. This paper proposes a novel neuro-fuzzy decision modelling approach, using neural networks for automatically determining the key clinical characteristics influencing a physician's treatment decisions, and using fuzzy classifiers to model the relationships between clinical characteristics and treatment decisions using linguistically interpretable fuzzy rules. The approach aims to help identify the factors and reasons for variations in treatment decisions made by different physicians in order to improve patient care. In order to demonstrate the usefulness of the proposed work, we conducted several quick and dirty ethnographic studies, which prove that variations in physicians' treatment exist.
- Oct 2011
This paper presents the development of a neuro-fuzzy agent for ambient-intelligence environments. The agent has been implemented as a system-on-chip (SoC) on a reconfigurable device, i.e., a field-programmable gate array. It is a hardware/software (HW/SW) architecture developed around a MicroBlaze processor (SW partition) and a set of parallel intellectual property cores for neuro-fuzzy modeling (HW partition). The SoC is an autonomous electronic device able to perform real-time control of the environment in a personalized and adaptive way, anticipating the desires and needs of its inhabitants. The scheme used to model the intelligent agent is a particular class of an adaptive neuro-fuzzy inference system with piecewise multilinear behavior. The main characteristics of our model are computational efficiency, scalability, and universal approximation capability. Several online experiments have been performed with data obtained in a real ubiquitous computing environment test bed. Results obtained show that the SoC is able to provide high-performance control and adaptation in a life-long mode while retaining the modeling capabilities of similar agent-based approaches implemented on larger computing machines.
In this paper an ambient intelligent-based framework is proposed for the monitoring of dementia patients living in their own homes. Within this framework groups of unobtrusive wireless sensor devices can be deployed at specific locations within a patient's home and accessed via standardized interfaces provided through an open middleware platform. For each sensor group intelligent agents are used to learn fuzzy rules, which model the patient's habitual behaviours in the environment. An online rule adaptation technique is applied to facilitate short-term tuning of the learnt behaviours, and long-term tracking of behaviour changes which could be due to the effects of cognitive decline caused from dementia. The proposed system reports behaviour changes to care providers to enable them to make better-informed assessments of the patient's cognitive abilities and changing care needs.
- Oct 2010
The evidence suggests that human actions are supported by emotional elements that complement logic inference in our decision-making processes. In this paper an exploratory study is presented providing initial evidence of the positive effects of emotional information on the ability of intelligent agents to create better models of user actions inside smart-homes. Preliminary results suggest that an agent incorporating valence-based emotional data into its input array can model user behaviour in a more accurate way than agents using no emotion-based data or raw data based on physiological changes.
‘iSpace, the final frontier’ — this parody of Star Trek encapsulates many of our aspirations for this area as, in the longer term, iSpaces are likely to be the key to mankind’s successful exploration of deep space. In outer space, or hostile planetary habitats, it is inevitable that people will survive in wholly technologically supported artificial environments . Such environments will contain numerous communicating computers embedded into a myriad of devices, sensing, acting, delivering media, processing data, and providing services that enhance the life-style and effectiveness of the occupant and, in outer space, preserving human life. Such environments will also include robots . In today’s iSpaces, while human life will not normally be at stake, the underlying principles and technology are much the same. Today our homes are rapidly being filled with diverse types of products ranging from simple lighting systems to sophisticated entertainment systems, all adding to the functionality and convenience available to the home user. The iSpace approach envisages that, one day soon, most artefacts will contain embedded computers and network connections, opening up the possibility for hundreds of communicating devices, co-operating in communities serving the occupant(s). The seeds of this revolution have already been sown in that pervasive technologies such as the Internet and mobile telephones already boast over 200 and 680 million users, respectively .
Applicant selection and ranking methods for job roles within human resources (HR) systems involve high levels of uncertainty. This is due to the requirement to allow for the varying opinions and preferences of the different occupation domain experts in the decision making process. Hence, there is a need to develop novel systems that will enable HR departments to determine the most important requirements criteria (experience, skills etc) for a given job, based on the preferences of different domain experts, while ensuring that the experts decisions are unbiased and correctly weighted according to their knowledge and experience. This will enable a more effective way to short list submitted candidate CVs from a large number of applicants providing a consistent and fair CV ranking policy, which can be legally justified. This paper presents a novel system using a neuro-fuzzy based agent approach for automatically determining the key skill characteristics defining each expert's preferences and ranking decisions, while handling the uncertainties and inconsistencies in group decisions of a panel of experts. The presented system automates the processes of requirements specification and applicant's ranking. Experiments have been performed within the residential care sector where the proposed system has been shown to produce ranking decisions that were relatively highly consistent with those of the human experts.
- May 2009
- Intelligent Agents, 2009. IA '09. IEEE Symposium on
An effective applicant selection procedure for job roles is one of the most significant requirements for organisations human resources (HR) departments. Due to the high number of applicants it is necessary to short-list and rank submitted CVs based on their suitability for the job requirements. To reduce costs, error and time there is a strong desire from companies towards automating the two processes of: specifying the requirements criteria for a given job (experience, skills, etc) and matching between the applicants' profiles and the job requirements; to produce an applicants' ranking policy that gives consistent and fair results which can be legally justified. However both these processes involve a high level of uncertainty, as they require the input of different occupation domain experts in the decision making process. These experts will have different opinions, expectations and interpretations for the requirements specification as well as for the applicants matching and ranking criteria. Determining the consistency and reliability of each expert's decision making behaviours is also necessary to ensure that experts decisions are unbiased and correctly weighted according to their level knowledge and experience. This paper presents a novel approach for ranking job applicants by employing fuzzy agents for handling the uncertainties and inconsistencies in group decisions of a panel of experts. The presented system will enable automating the processes of requirements specification and applicant's matching/ranking. Experiments have been performed within the residential care sector in which the proposed system has been shown to produce ranking decisions that were relatively highly consistent with those of the human experts.
- Jul 2008
- Fuzzy Systems, 2008. FUZZ-IEEE 2008. (IEEE World Congress on Computational Intelligence). IEEE International Conference on
Global warming is becoming one of the serious issues facing humanity. Several initiatives have been introduced to deal with global warming including the Kyoto protocol which assigned mandatory targets for the reduction of greenhouse gas emissions to signatory nations. However, over the last decade, commercial buildings worldwide have experienced massive growth in energy costs. This was caused by the expansion in the use of air conditioning and artificial lighting as well as an ever increasing energy demand for computing services. Existing building management systems (BMSs) have, generally, failed to fully optimize energy consumption in commercial buildings. This is because they lack control systems that can react intelligently and automatically to anticipated changes in ambient weather conditions and the many other environmental variables typically associated with large buildings. In this paper, we present a novel agent based system entitled intelligent control of energy (ICE) for energy management in commercial buildings. ICE uses different computational intelligence (CI) techniques (including fuzzy systems, neural networks and genetic algorithms) to dasialearnpsila a buildings thermal response to many variables including the outside weather conditions, internal occupancy requirements and building plant responses. ICE then uses CI based algorithms which work in real-time with the buildingpsilas existing BMS to minimize the buildingpsilas energy demand. We will show how the use of ICE will allow significant energy cost savings, while still maintaining customer-defined comfort levels.
Ranking applicants for a given job is one of the most important processes for Human Resources (HR) systems. The ranking of job applicants involves two main processes which are the specification of the requirements criteria for a given job (experience, skills, etc) and the matching between the applicantspsila profiles and the job requirements. There is currently a strong move towards automating these two processes to generate an applicantspsila ranking system that gives consistent and fair results. However there is a high level of uncertainty involved in these two processes as they involve the input of several experts. These experts will have different opinions, expectations the interpretations for the requirements specification as well as for the applicants matching and ranking. This paper presents a novel approach for ranking job applicants by employing type-2 fuzzy sets for handling the uncertainties in group decisions in a panel of experts. Hence the presented system will enable automating the processes of requirements specification and applicants matching/ranking. We have performed real world experiments in the care domain where our system handled the uncertainties and produced ranking decisions that were consistent with those of the human experts. To the authorspsila knowledge, this will be the first type-2 based commercial software system.
Ambient Intelligence is nowadays an active research field. As a key part of this concept, learning architectures for the control of the devices in an intelligent building must be developed, where the goal is to control the environment via a set of devices using an intelligent agent which should work in a non-intrusive manner to satisfy the preferences of the user. Mainly, we have focused our attention over fuzzy logic controllers (FLC) for the internal structure of the agent. The main motivation for the work described in this paper is to check different alternatives in order to select a suitable method for the off-line data driven automatic generation of FLC for the agent. We have performed the experiments with real data gathered from the Essex Intelligent Dormitory.
In this paper, we present a novel type-2 fuzzy systems based adaptive architecture for agents embedded in ambient intelligent environments (AIEs). Type-2 fuzzy systems are able to handle the different sources of uncertainty and imprecision encountered in AIEs to give a very good response. The presented agent architecture uses a one pass method to learn in a nonintrusive manner the user's particular behaviors and preferences for controlling the AIE. The agent learns the user's behavior by learning his particular rules and interval type-2 Membership Functions (MFs), these rules and MFs can then be adapted online incrementally in a lifelong learning mode to suit the changing environmental conditions and user preferences. We will show that the type-2 agents generated by our one pass learning technique outperforms those generated by genetic algorithms (GAs). We will present unique experiments carried out by different users over the course of the year in the Essex Intelligent Dormitory (iDorm), which is a real AIE test bed. We will show how the type-2 agents learnt and adapted to the occupant's behavior whilst handling the encountered short term and long term uncertainties to give a very good performance that outperformed the type-1 agents while using smaller rule bases
In this paper, we will present a novel system for learning and incrementally adapting type-2 Fuzzy Logic Controllers (FLCs) for agents embedded in Ambient Intelligent Environments (AIEs). The system learns the rules and the type-2 Membership Functions (MFs) for the type-2 FLC that models the user behavior. Over long term operations, the agent incrementally adapts the type-2 FLC rules and MFs in a life long learning mode to accommodate for the short term and long term uncertainties encountered in AIEs. We will present unique experiments carried out by different users over the course of the year in the Essex intelligent Dormitory (iDorm) which is a real AIE test bed. We will show how the type-2 agent learnt and adapted to the occupant's behavior, whilst handling the encountered short term and long term uncertainties to give a very good performance that outperformed the type-1 fuzzy agents while using smaller rule bases.
- Jul 2005
- Intelligent Environments, 2005. The IEE International Workshop on (Ref. No. 2005/11059)
Ambient Intelligence is nowadays an active research field. As a key matter of this concept, learning architectures for the control of the devices in an intelligent building must be developed, where the goal is to control the environmental via a set of devices using an intelligent agent which should work in a non-intrusive manner to satisfy the preferences of the user. Mainly, we have focused our attention over fuzzy logic controllers (FLC) for the internal structure of the agent. The main motivation for the work described in this paper is to check different alternatives selecting a suitable method for the off-line data driven automatic generation of FLCs for the agent. We have performed the experiments with real data gathered from the Essex Intelligent Dormitory.
- May 2005
In this paper, we present a novel approach for realising the vision of ambient intelligence in ubiquitous computing environments (UCEs). This approach is based on embedding intelligent agents in UCEs. These agents use type-2 fuzzy systems which are able to handle the different sources of uncertainty and imprecision in UCEs to give a good response. We have developed a novel system for learning and adapting the type-2 fuzzy agents so that they can realise the vision of ambient intelligence by providing a seamless, unobtrusive, adaptive and responsive intelligence in the environment that supports the activities of the user. The user’s behaviours and preferences for controlling the UCE are learnt online in a non-intrusive and life long learning mode so as to control the UCE on the user’s behalf. We have performed unique experiments in which the type-2 intelligent agent has learnt and adapted online to the user’s behaviour during a stay of five days in the intelligent Dormitory (iDorm) which is a real UCE test bed. We will show how our type-2 agents can deal with the uncertainty and imprecision present in UCEs to give a very good response that outperforms the type-1 fuzzy agents while using smaller rule bases.
We describe a novel life-long learning approach for intelligent agents that are embedded in intelligent environments. The agents aim to realize the vision of ambient intelligence in intelligent inhabited environments (IIE) by providing ubiquitous computing intelligence in the environment supporting the activities of the user. An unsupervised, data-driven, fuzzy technique is proposed for extracting fuzzy membership functions and rules that represent the user's particularized behaviors in the environment. The user's learned behaviors can then be adapted online in a life-long mode to satisfy the different user and system objectives. We have performed unique experiments in which the intelligent agent has learned and adapted to the user's behavior, during a stay of five consecutive days in the intelligent dormitory (iDorm), which is a real ubiquitous computing environment test bed. Both offline and online experimental results are presented comparing the performance of our technique with other approaches. The results show that our proposed system has outperformed the other approaches, while operating online in a life-long mode to realize the ambient intelligence vision.
- Oct 2004
- Systems, Man and Cybernetics, 2004 IEEE International Conference on
Ambient intelligence is nowadays an active research field. As a key matter of this concept, several approaches have been proposed for the development of learning architectures for the control of the devices in an intelligent building. In this paper, an evolutionary algorithm is analyzed as a candidate for the initial phases of the design of such architectures: fuzzy controllers for the devices are offline induced from data sampled from the environment. We would show results obtained using real data gathered from the Essex intelligent dormitory. The proposed algorithm seems to be suited for the task, both due to its accuracy and for the easy and meaningful linguistic interpretation of the solutions it produces.
We describe a novel system for learning and adapting type-2 fuzzy controllers for intelligent agents that are embedded in ubiquitous computing environments (UCEs). Our type-2 agents operate non intrusively in an online life long learning manner to learn the user behaviour so as to control the UCE on the user's behalf. We have performed unique experiments in which the type-2 intelligent agent has learnt and adapted online to the user's behaviour during a stay of five days in the intelligent dormitory (iDorm) which is a real UCE test bed. We show how our type-2 agent deals with the uncertainty and imprecision present in UCEs to give a very good performance that outperform the type-1 fuzzy agents while using a smaller number of rules.
There is an increasing amount of research into the area of pervasive computing, smart homes and intelligent spaces, one example being that of the DTI-funded Pervasive Home Environment Networking (PHEN) project. Much of the current research focuses on environments populated by numerous computing devices, sensors, actuators, various wired and wireless networking systems and poses the question of how such computing environments might become intelligent? Often, the proposed solution is to explicitly preprogram in the intelligence. In this paper we discuss a solution based on embedded-agents which enables emergent intelligent behaviour by predominantly implicit processes. We describe an experimental test-bed for pervasive computing, the iDorm, and report on experiments that scope the agent-learning characteristics in such environments. We also introduce a more human-directed approach to programming in pervasive environments which we refer to as task-oriented programming (TOP).
In this paper we describe a novel life long learning approach for intelligent agents that are embedded in intelligent environments. The agents aim to realise the vision of Ambient Intelligence in Intelligent Inhabited Environments (IIE) by providing 'ubiquitous computing intelligence in the environment supporting the activities of the user. An unsupervised, data-driven, fuzzy, technique is proposed for extracting fuzzy membership functions and rules that represent the user's particularised behaviours in the environment. The user's learnt behaviours can then be adapted online in a life long mode to satisfy the different user and system objectives. We have performed unique experiments in which the intelligent agent has learnt and adapted to the user's behaviour, during a stay of five consecutive days in the intelligent Dormitory (iDorm) which is a real ubiquitous computing environment test bed. Both offline and online experimental results are presented comparing the performance of our technique with other approaches. The results show that our proposed system has outperformed the other systems while operating online in a life long mode to realise the ambient intelligence vision.
Ambient Intelligence (AiM) is nowadays an active research field. As a key matter of this concept, several approaches have been proposed for the development of learning techniques for the control of the devices in an intelligent building. In this paper, a GA-P is analyzed and discussed as a candidate algorithm for the design of such learning architectures. Some results obtained from experiments using real data gathered from an intelligent dormitory are shown. The GA-P seems to be a suitable method for the task, both due to its accuracy and for the easy and meaningful linguistic interpretation of the solutions it produces. A comparison with other learning method (Anfis) is also included.
Resumen— Recientemente, la investigación en el campo de la Inteligencia Ambiental está tomando gran auge en la comunidad científica. Dentro de este paradigma, el estado del arte recoge diferen-tes intentos de aplicación de técnicas de aprendi-zaje al control de los dispositivos de una vivienda en base a un aprendizaje no intrusivo realizado a partir de muestras de datos de diferentes senso-res y del estado de los dispositivos de la vivienda. En este artículo se analizan las posibilidades de aplicación de algoritmos GA-P al diseño de ta-les arquitecturas de aprendizaje, particularizando la implementación a la obtención de controlado-res difusos de los dispositivos de la vivienda. Se muestran algunos resultados iniciales obtenidos a partir de datos reales tomados en un dormitorio inteligente. El algoritmo GA-P propuesto se mues-tra como un candidato adecuado en taí ambito de aplicación, tanto desde el punto de vista de la precisión conseguida en el control de los disposi-tivos como en lo referente a la interpretabilidad lingüística de las soluciones.