Article

Gender recognition using optimal gait feature based on recursive feature elimination in normal walking

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

This study aims to propose a novel approach for gender recognition using best feature subset based on recursive feature elimination (RFE) in normal walking. This study has focused on the analysis of gait characteristics by distinguishing the gait phases as initial contact (IC), Mid-stance (MS), Pre-swing, and swing (SW), and collected the large number of gait to improve the reliability of quantitative assessment of natural variability associated with muscle activity during free walking. The gait system was designed using pressure and a tri-axis accelerometer sensor, and a 9-channel electromyography sensor for measuring the data. Gender recognition method was proposed using support vector machine (SVM) and random forest (RF) based on RFE to determine best feature subset. Statistical results show that effects of gender-based differences on gait characteristic including temporal, kinematics, and muscle activity were investigated. The temporal parameters of stride time and gait cycle (%) in the gait phases of IC, MS, and SW were significantly different between females and males (p<0.01). The females exhibited both a lower angle and a root mean square acceleration of the knee joint as compared to the males, and there was a clear gender-based difference with respect to knee angle movement. In addition, most muscle activation measurements in the females were larger than those of the males with respect to the gait phases. Gender classification result shows that SVM-RFE was 99.11% (SVM classifier) and RF-RFE was 98.89% (SVM and RF classifier), having powerful performance.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... But it is not always the same viewpoint in real-time. Unfortunately, among the recent research literature [13][14][15][16][17][18][25][26][27][28][29], very few have conducted gait-based gender classification experiments for viewing angle variations on a very large dataset, which can be considered as the main motivation of this study. ...
... However, in other viewing angles, the subjects' limb movement can be occluded from the camera, which hinders perceiving the gait features. Recent studies [18,[28][29][30] have tried to investigate the problem of viewing angle variations in gait-based gender classification. However, they have used a relatively smaller dataset for performance evaluation of the results. ...
... Further, the features with significant differences by gender were used to train a support vector machine (SVM) classifier. Lee et al. [29] proposed a gender recognition method that uses a support vector machine (SVM) and random forest (RF) based on recursive feature elimination to determine the best features. They investigated temporal, kinematics, and muscle activity to show the effects of gender-based differences on gait characteristics. ...
Article
Full-text available
In computer vision applications, gait-based gender classification is a challenging task as a person may walk at various angles with respect to the camera viewpoint. In some of the viewing angles, the person’s limb movement can be occluded from the camera, preventing the perception of the gait-based features. To solve this problem, this study proposes a robust and lightweight system for gait-based gender classification. It uses a gait energy image (GEI) for representing the gait of an individual. A discrete cosine transform (DCT) is applied on GEI to generate a gait-based feature vector. Further, this DCT feature vector is applied to XGBoost classifier for performing gender classification. To improve the classification results, the XGBoost parameters are tuned. Finally, the results are compared with the other state-of-the-art approaches. The performance of the proposed system is evaluated on the OU-MVLP dataset. The experiment results show a mean CCR (correct classification rate) of 95.33% for the gender classification. The results obtained from various viewpoints of OU-MVLP illustrate the robustness of the proposed system for gait-based gender classification.
... Sex recognition has been studied using body [5] and face [6] images, gait [7], voice [8], twittering [9], and keystroke dynamics [10], among others. ...
Article
Full-text available
Physiological signals are strongly related to a person’s state of health and carry information about the human body. For example, by ECG, it is possible to obtain information about cardiac disease, emotions, personal identification, and the sex of a person, among others. This paper proposes the study of the heartbeat from a soft-biometric perspective to be applied to smartphone unlocking services. We employ the user heartbeat to classify the individual by sex (male, female) with the use of Deep Learning, reaching an accuracy of 94.4% ± 2.0%. This result was obtained with the RGB representation of the union of the time-frequency transformation from the pseudo-orthogonal X, Y, and Z bipolar signals. Evaluating each bipolar contribution, we found that the XYZ combination provides the best category distinction using GoogLeNet. The 24-h Holter database of the study contains 202 subjects with a female size of 49.5%. We propose an architecture for managing this signal that allows the use of a few samples to train the network. Due to the hidden nature of ECG, it does not present vulnerabilities like public trait exposition, light/noise sensibility, or learnability compared to fingerprint, facial, voice, or password verification methods. ECG may complement those gaps en route to a cooperative authentication ecosystem.
... Multiple feature selection techniques such as a wrapper, and embedded and filter-based techniques are widely utilized to obtain a set of important features. Recursive feature elimination is one of the most important techniques for feature selection which involves eliminating features at each level and re-ranking the rest of the features by retraining the existing features (Lee et al., 2022). Figure 5 shows the working principle of the recursive feature elimination approach. ...
Article
The prediction of lifestyle diseases is a vital domain in healthcare informatics research. This task is primarily achieved using the widely available machine learning algorithms. However, the high-dimensionality of data amplifies the computation complexity and significantly reduces the models' efficiency. Conspicuously, we presented a multi-disease prediction strategy for intelligent decision support using ensemble learning. The proposed work leverages genetic algorithm-based recursive feature elimination and AdaBoost to predict two prominent lifestyle diseases. Alongside the proposed approach, various benchmark machine learning techniques are also trained and validated using selected features under k-fold cross-validation. The results reveal the effectiveness of the proposed methodology in predicting multiple diseases in comparison to past works.
... Feature extraction and feature selection (feature dimension reduction) are not completely separated in practical application; there are few studies on the combination of the two and further research is needed [54]. Finally, the purpose of this research was to build a saline-alkali rice variety recognition model using stable and efficient machine learning algorithms, the literature has confirmed the superior performance of RFM in classification evaluation [53,55]; the RFE method can eliminate the features of a low contribution rate without increasing model errors through repeated iterations to select the best features [33,56]. Therefore, RFE dimension reduction method was combined with the RFM in this study to improve the recognition rate of the identification model. ...
Article
Full-text available
Planting rice in saline–alkali land can effectively improve saline–alkali soil and increase grain yield, but traditional identification methods for saline–alkali-tolerant rice varieties require tedious and time-consuming field investigations based on growth indicators by rice breeders. In this study, the Python machine deep learning method was used to analyze the Raman molecular spectroscopy of rice and assist in feature attribution, in order to study a fast and efficient identification method of saline–alkali-tolerant rice varieties. A total of 156 Raman spectra of four rice varieties (two saline–alkali-tolerant rice varieties and two saline–alkali-sensitive rice varieties) were analyzed, and the wave crests were extracted by an improved signal filtering difference method and the feature information of the wave crest was automatically extracted by scipy.signal.find_peaks. Select K Best (SKB), Recursive Feature Elimination (RFE) and Select F Model (SFM) were used to select useful molecular features. Based on these feature selection methods, a Logistic Regression Model (LRM) and Random Forests Model (RFM) were established for discriminant analysis. The experimental results showed that the RFM identification model based on the RFE method reached a higher recognition rate of 89.36%. According to the identification results of RFM and the identification of feature attribution materials, amylum was the most significant substance in the identification of saline–alkali-tolerant rice varieties. Therefore, an intelligent method for the identification of saline–alkali-tolerant rice varieties based on Raman molecular spectroscopy is proposed.
... Automatic gender recognition has grown critical in recent years, particularly in crime detection. Numerous modalities have been employed to identify female and male subjects, including the face (Hsu et al. 2021), voice (Livieris et al. 2019), gait (Lee et al. 2022), hand (Afifi 2019), fingerprint (Jalali et al. 2021), and electrocardiogram (ECG) signals (Uçar et al. 2021). ...
Article
Full-text available
Human facial analysis (HFA) has recently become an attractive topic for computer vision research due to technological progress and mobile applications. HFA explores several issues as gender recognition (GR), facial expression, age, and race recognition for automatically understanding social life. This study explores HFA from the angle of recognizing a person’s gender from their face. Several hard challenges are provoked, such as illumination, occlusion, facial emotions, quality, and angle of capture by cameras, making gender recognition more difficult for machines. The Archimedes optimization algorithm (AOA) was recently designed as a metaheuristic-based population optimization method, inspired by the Archimedes theory’s physical notion. Compared to other swarm algorithms in the realm of optimization, this method promotes a good balance between exploration and exploitation. The convergence area is increased By incorporating extra data into the solution, such as volume and density. Because of the preceding benefits of AOA and the fact that it has not been used to choose the best area of the face, we propose utilizing a wrapper feature selection technique, which is a real motivation in the field of computer vision and machine learning. The paper’s primary purpose is to automatically determine the optimal face area using AOA to recognize the gender of a human person categorized by two classes (Men and women). In this paper, the facial image is divided into several subregions (blocks), where each area provides a vector of characteristics using one method from handcrafted techniques as the local binary pattern (LBP), histogram-oriented gradient (HOG), or gray-level co-occurrence matrix (GLCM). Two experiments assess the proposed method (AOA): The first employs two benchmarking datasets: the Georgia Tech Face dataset (GT) and the Brazilian FEI dataset. The second experiment represents a more challenging large dataset that uses Gallagher’s uncontrolled dataset. The experimental results show the good performance of AOA compared to other recent and competitive optimizers for all datasets. In terms of accuracy, the AOA-based LBP outperforms the state-of-the-art deep convolutional neural network (CNN) with 96.08% for the Gallagher’s dataset.
Article
Full-text available
The increasing amount of data in electromyographic (EMG) signal research has greatly increased the importance of developing advanced data analysis and machine learning techniques which are better able to handle "big data". Consequently, more advanced applications of EMG pattern recognition have been developed. This paper begins with a brief introduction to the main factors that expand EMG data resources into the era of big data, followed by the recent progress of existing shared EMG data sets. Next, we provide a review of recent research and development in EMG pattern recognition methods that can be applied to big data analytics. These modern EMG signal analysis methods can be divided into two main categories: (1) methods based on feature engineering involving a promising big data exploration tool called topological data analysis; and (2) methods based on feature learning with a special emphasis on "deep learning". Finally, directions for future research in EMG pattern recognition are outlined and discussed.
Article
Full-text available
Routine stress monitoring in daily life can predict potentially serious health impacts. Effective stress monitoring in medical and healthcare fields is dependent upon accurate determination of stress-related features. In this study, we determined the optimal stress-related features for effective monitoring of cumulative stress. We first investigated the effects of short- and long-term stress on various heart rate variability (HRV) features using a rodent model. Subsequently, we determined an optimal HRV feature set using support vector machine-recursive feature elimination (SVM-RFE). Experimental results indicate that the HRV time domain features generally decrease under long-term stress, and the HRV frequency domain features have substantially significant differences under short-term stress. Further, an SVM classifier with a radial basis function kernel proved most accurate (93.11%) when using an optimal HRV feature set comprising the mean of R-R intervals (mRR), the standard deviation of R-R intervals (SDRR), and the coefficient of variance of R-R intervals (CVRR) as time domain features, and the normalized low frequency (nLF) and the normalized high frequency (nHF) as frequency domain features. Our findings indicate that the optimal HRV features identified in this study can effectively and efficiently detect stress. This knowledge facilitates development of in-facility and mobile healthcare system designs to support stress monitoring in daily life.
Article
Full-text available
Feature selection, which identifies a set of most informative features from the original feature space, has been widely used to simplify the predictor. Recursive feature elimination (RFE), as one of the most popular feature selection approaches, is effective in data dimension reduction and efficiency increase. A ranking of features, as well as candidate subsets with the corresponding accuracy, is produced through RFE. The subset with highest accuracy (HA) or a preset number of features (PreNum) are often used as the final subset. However, this may lead to a large number of features being selected, or if there is no prior knowledge about this preset number, it is often ambiguous and subjective regarding final subset selection. A proper decision variant is in high demand to automatically determine the optimal subset. In this study, we conduct pioneering work to explore the decision variant after obtaining a list of candidate subsets from RFE. We provide a detailed analysis and comparison of several decision variants to automatically select the optimal feature subset. Random forest (RF)-recursive feature elimination (RF-RFE) algorithm and a voting strategy are introduced. We validated the variants on two totally different molecular biology datasets, one for a toxicogenomic study and the other one for protein sequence analysis. The study provides an automated way to determine the optimal feature subset when using RF-RFE.
Article
Full-text available
Feature selection is an important topic in bioinformatics. Defining informative features from complex high dimensional biological data is critical in disease study, drug development, etc. Support vector machine-recursive feature elimination (SVM-RFE) is an efficient feature selection technique that has shown its power in many applications. It ranks the features according to the recursive feature deletion sequence based on SVM. In this study, we propose a method, SVM-RFE-OA, which combines the classification accuracy rate and the average overlapping ratio of the samples to determine the number of features to be selected from the feature rank of SVM-RFE. Meanwhile, to measure the feature weights more accurately, we propose a modified SVM-RFE-OA (M-SVM-RFE-OA) algorithm that temporally screens out the samples lying in a heavy overlapping area in each iteration. The experiments on the eight public biological datasets show that the discriminative ability of the feature subset could be measured more accurately by combining the classification accuracy rate with the average overlapping degree of the samples compared with using the classification accuracy rate alone, and shielding the samples in the overlapping area made the calculation of the feature weights more stable and accurate. The methods proposed in this study can also be used with other RFE techniques to define potential biomarkers from big biological data.
Thesis
Full-text available
Muscles are the key drivers in any human movement. Since the muscles generate the forces and consequently the impulses to move the athlete from one position to another, it can be useful to study the muscle activity during sports movements to help with optimisation of technique, injury prevention and performance enhancements. Due to recent advances in electromyography (EMG) technologies, muscle activity in sports movement such as shot putting and overground sprinting can now be acquired using wireless surface mount sensors. Previously the use of tethered devices restricted the movements which could be analysed. The aim of this research was to investigate data analysis methods for use with EMG. There is a need to develop an in depth understanding of what EMG data can convey by understanding muscle activations and patterns in various sports movements and techniques. The research has been implemented by conducting a literature review, a survey and experimental studies to examine EMG signals on shot putting, sprinting and to understand cross-talk. There has been significant work done in understanding the biomechanics of sprinting, with emphasis on kinematics. The literature review on muscle activities in sprinting highlighted the need for wireless devices to allow testing of athletes in ecologically valid environments, rather than on a treadmill which offers little comparison with the environment of the sprinter, and proposed that there existed a bias on the muscles studied which may have been due to technology constraints of tethered systems. The survey of biomechanists gave an insight into the sensor devices utilised, the types of experimental studies being undertaken and the specifications desired in these devices. The study of muscle activations during the glide technique in shot put delivered meaningful activation patterns which coincided with key movements in the technique and augmented previously known kinematic data and anecdotal evidence. The study on muscle activations during maximal sprinting returned similar results, the 50% threshold provided information on the higher volume of muscle activity and these bursts of activity also coincided with key kinematic events. The use of independent component analysis (ICA) was examined to reduce cross-talk during sporting movements and recreating EMG signals due incorrectly positioned electrodes. Few studies have examined ICA with myoelectric signals. This research applies ICA to EMG signals during isometric contractions; small increases in correlation were found in some cases between the output signals and the ideal signals. The data analysis methods used in this research along with the supporting studies may prove to be a vital aid in supporting practitioners, coaches and athletes in the analysis of shot putting and sprinting using muscle activations and patterns. The thresholding methods used in this work may be useful in future studies to distinguish between low and high volumes of EMG activity in sports movements. It is recommended that future studies examine the muscle activity of specific exercises and compare the activity to that of sports movements to determine which exercises are most suitable in training and for pre-activation. The ICA algorithm should be examined further, to analyse isotonic movements.
Article
Full-text available
Gender recognition is trivial for a physiotherapist, but it is considered a challenge for computers. The electromyography (EMG) and heart rate variability (HRV) were utilized in this work for gender recognition during exercise using a stepper. The relevant features were extracted and selected. The selected features were then fused to automatically predict gender recognition. However, the feature selection for gender classification became a challenge to ensure better accuracy. Thus, in this paper, a feature selection approach based on both the performance and the diversity between the two features from the rank-score characteristic (RSC) function in a combinatorial fusion approach (CFA) (Hsu et al.) was employed. Then, the features from the selected feature sets were fused using a CFA. The results were then compared with other fusion techniques such as naive bayes (NB), decision tree (J48), k-nearest neighbor (KNN) and support vector machine (SVM). Besides, the results were also compared with previous researches in gender recognition. The experimental results showed that the CFA was efficient and effective for feature selection. The fusion method was also able to improve the accuracy of the gender recognition rate. The CFA provides much better gender classification results which is 94.51% compared to Barani's work (90.34%), Nazarloo's work (92.50%), and other classifiers.
Article
Full-text available
This research article attempts to provide a recent survey on neuro-fuzzy approaches for feature selection and classification. Feature selection acts as a catalyst in reducing computation time and dimensionality, enhancing prediction performance or accuracy and curtailing irrelevant or redundant data. The neuro-fuzzy approach is used for feature selection and for providing some insight to the user about the symbolic knowledge embedded within the network. The neuro–fuzzy approach combines the merits of neural network and fuzzy logic to solve many complex machine learning problems. The objective of this article is to provide a generic introduction and a recent survey to neuro-fuzzy approaches for feature selection and classification in a wide area of machine learning problems. Some of the existing neuro-fuzzy models are also applied on standard datasets to demonstrate the applicability of neuro-fuzzy approaches.
Conference Paper
Full-text available
Electromyography (EMG) is an experiment-based method for evaluating and recording a series of electrical signals that emanate from body muscles. The electrical manifestation of neuromuscular activation generated in muscles during contraction and/or relaxation is known as EMG signals. In this paper, a preliminary study is conducted in order to improve the fitness of post-stroke survivors with a minimal supervision from therapists in physiological activity especially on the lower limb rehabilitation. Therefore, a pattern recognition technique is required to extract the important features of an EMG signal to control the physiological devices (PDs), for instance, cycling-like and stepping-like machines in a lower limb rehab application. A new approach for feature extraction vectors in a recognition system will be proposed using Discrete Wavelet Transform (DWT) and Fuzzy C-Means (FCM) algorithms. In addition to this, a Principle Component Analysis (PCA) method will be utilized to reduce the dimension of data in prior to computing the classification accuracy using the Adaptive Neuro-Fuzzy Inference System (ANFIS).
Article
Full-text available
This paper presents a fast and accurate patient-specific electrocardiogram (ECG) classification and monitoring system. An adaptive implementation of 1D Convolutional Neural Networks (CNNs) is inherently used to fuse the two major blocks of the ECG classification into a single learning body: feature extraction and classification. Therefore, for each patient an individual and simple CNN will be trained by using relatively small common and patient-specific training data, and thus such a patient-specific feature extraction ability can further improve the classification performance. Since this also negates the necessity to extract hand-crafted manual features, once a dedicated CNN is trained for a particular patient, it can solely be used to classify possibly long ECG data stream in a fast and accurate manner or alternatively, such a solution can conveniently be used for real-time ECG monitoring and early alert system on a light-weight wearable device. The results over the MIT-BIH arrhythmia benchmark database demonstrate that the proposed solution achieves a superior classification performance than most of the state-of-the-art methods for the detection of ventricular ectopic beats (VEB) and supraventricular ectopic beats (SVEB). Besides the speed and computational efficiency achieved, once a dedicated CNN is trained for an individual patient, it can solely be used to classify his/her long ECG records such as Holter registers in a fast and accurate manner. Due to its simple and parameter invariant nature, the proposed system is highly generic and thus applicable to any ECG dataset.
Article
Full-text available
Support vector machine recursive feature elimination (SVM-RFE) is a powerful feature selection algorithm. However, when the candidate feature set contains highly correlated features, the ranking criterion of SVM-RFE will be biased, which would hinder the application of SVM-RFE on gas sensor data. In this paper, the linear and nonlinear SVM-RFE algorithms are studied. After investigating the correlation bias, an improved algorithm SVM-RFE + CBR is proposed by incorporating the correlation bias reduction (CBR) strategy into the feature elimination procedure. Experiments are conducted on a synthetic dataset and two breath analysis datasets, one of which contains temperature modulated sensors. Large and comprehensive sets of transient features are extracted from the sensor responses. The classification accuracy with feature selection proves the efficacy of the proposed SVM-RFE + CBR. It outperforms the original SVM-RFE and other typical algorithms. An ensemble method is further studied to improve the stability of the proposed method. By statistically analyzing the features’ rankings, some knowledge is obtained, which can guide future design of e-noses and feature extraction algorithms.
Article
Full-text available
The selection of optimal features has long been a subject of debate for pattern recognition based myoelectric control. Studies have compared many features, but have typically used small or constrained data sets. Herein, the performance of various features is evaluated using data from six previously reported data sets. The number of channels, the contraction dynamics (dynamic vs static), and classifier type all yielded significant interactions (p<;0.05) with the feature set. When using 8 channels, the addition of the tested features produced no improvement over a standard time domain (TD) feature set alone (p>0.05). When using fewer channels, however, autoregressive, Cepstral coefficients, Willison amplitude and sample entropy features all provided significant improvement during dynamic contractions (p<;0.05). The simple Willison amplitude is highlighted, showing that it can provide significant improvement when used as a replacement for any one of the standard TD features.
Article
Full-text available
Electromyographic (EMG) signals have been widely employed as a control signal in rehabilitation and a means of diagnosis in health care. Signal amplification and filtering is the first step in surface EMG signal processing and application systems. The characteristics of the amplifiers and filters determine the quality of EMG signals. Up until now, searching for better amplification and filtering circuit design that is able to accurately capture the features of surface EMG signals for the intended applications is still a challenging. With the fast development in computer sciences and technologies, EMG signals are expected to be used and integrated within small or even tiny intelligent, automatic, robotic, and mechatronics systems. This research focused on small size amplification and filtering circuit design for processing surface EMG signals from an upper limb and aimed to fix the amplifiers and filters inside a robotic hand with limited space to command and control the robot hand movement. The research made a study on the commonly used methodologies for EMG signal processing and circuitry design and proposed a circuit design for EMG signal amplification and filtering. High-pass filters including second-order and fourth-order with the suppression to low frequency noises are studied. The analysis, verification and experiment showed that a second-order high-pass filter can adequately suppress the low frequency noises. The proposed amplification and filtering circuit design is able to effectively clean the noises and collect the useful surface EMG signals from an upper limb. The experiment also clearly revealed that power line interference needs to be carefully handled for higher signal-noise-ratio (SNR) as a notch-filter might cause the loss of useful signal components. Commercial computer software such as LabView and Matlab were used for data acquisition software development and data analysis.
Article
Full-text available
Early detection of ventricular fibrillation (VF) and rapid ventricular tachycardia (VT) is crucial for the success of the defibrillation therapy. A wide variety of detection algorithms have been proposed based on temporal, spectral, or complexity parameters extracted from the ECG. However, these algorithms are mostly constructed by considering each parameter individually. In this work, we present a novel life-threatening arrhythmias detection algorithm which combines a number of previously proposed ECG parameters by using support vector machines (SVM) classifiers. A total of 13 parameters were computed accounting for temporal (morphological), spectral, and complexity features of the ECG signal. A filter-type feature selection (FS) procedure was proposed to analyze the relevance of the computed parameters and how they affect the detection performance. The proposed methodology was evaluated in two different binary detection scenarios: shockable (FV plus VT) vs nonshockable arrhythmias, and VF vs nonVF rhythms, using the information contained in the MITDB, the CUDB, and the VFDB. Sensitivity (SE) and specificity (SP) analysis on the out of sample test data showed values of SE= 95%, SP= 99%, and SE= 92%, SP= 97% in the case of shockable and VF scenarios, respectively. Our algorithm was benchmarked against individual detection schemes, significantly improving their performance. Our results demonstrate that the combination of ECG parameters using statistical learning algorithms improves the efficiency for the detection of life-threatening arrhythmias.
Article
Full-text available
Human locomotor movements exhibit considerable variability and are highly complex in terms of both neural activation and biomechanical output. The building blocks with which the central nervous system constructs these motor patterns can be preserved in patients with various sensory-motor disorders. In particular, several studies highlighted a modular burst-like organization of the muscle activity. Here we review and discuss this issue with a particular emphasis on the various examples of adaptation of locomotor patterns in patients (with large fiber neuropathy, amputees, stroke and spinal cord injury). The results highlight plasticity and different solutions to reorganize muscle patterns in both peripheral and central nervous system lesions. The findings are discussed in a general context of compensatory gait mechanisms, spatiotemporal architecture and modularity of the locomotor program.
Article
Full-text available
Lower limb prostheses have traditionally been mechanically passive devices without electronic control systems. Microprocessor-controlled passive and powered devices have recently received much interest from the clinical and research communities. The control systems for these devices typically use finite-state-controllers to interpret data measured from mechanical sensors embedded within the prosthesis. In this paper we investigated a control system that relied on information extracted from myoelectric signals to control a lower limb prosthesis while amputee patients were seated. Sagittal plane motions of the knee and ankle can be accurately (>90%) recognized and controlled in both a virtual environment and on an actuated transfemoral prosthesis using only myoelectric signals measured from nine residual thigh muscles. Patients also demonstrated accurate (~90%) control of both the femoral and tibial rotation degrees of freedom within the virtual environment. A channel subset investigation was completed and the results showed that only five residual thigh muscles are required to achieve accurate control. This research is the first step in our long-term goal of implementing myoelectric control of lower limb prostheses during both weight-bearing and non-weight-bearing activities for individuals with transfemoral amputation.
Article
Full-text available
Smart phones are quite sophisticated and increasingly incorporate diverse and powerful sensors. One such sensor is the tri-axial accelerometer, which measures acceleration in all three spatial dimensions. The accelerometer was initially included for screen rotation and advanced game play, but can support other applications. In prior work we showed how the accelerometer could be used to identify and/or authenticate a smart phone user [11]. In this paper we extend that prior work to identify user traits such as sex, height, and weight, by building predictive models from labeled accelerometer data using supervised learning methods. The identification of such traits is often referred to as "soft biometrics" because these traits are not sufficiently distinctive or invariant to uniquely identify an individual---but they can be used in conjunction with other information for identification purposes. While our work can be used for biometric identification, our primary goal is to learn as much as possible about the smart phone user. This mined knowledge can be then be used for a number of purposes, such as marketing or making an application more intelligent (e.g., a fitness app could consider a user's weight when calculating calories burned).
Article
Full-text available
This study tackles the ECG classification problem by means of a methodology, which is able to enhance classification performance while simultaneously reducing the computational resources, making it specially adequate for its application in the improvement of ambulatory settings. For this purpose, the sequential forward floating search (SFFS) algorithm is applied with a new criterion function index based on linear discriminants. This criterion has been devised specifically to be a quality indicator in ECG arrhythmia classification. Based on this measure, a comprehensive feature set is analyzed with the SFFS algorithm, and the most suitable subset returned is additionally evaluated with a multilayer perceptron (MLP) to assess the robustness of the model. Aiming at obtaining meaningful estimates of the real-world performance and facilitating comparison with similar studies, the present contribution follows the Association for the Advancement of Medical Instrumentation standard EC57:1998 and the same interpatient division scheme used in several previous studies. Results show that by applying the proposed methods, the performance obtained in similar studies under the same constraints can be exceeded, while keeping the requirements suitable for ambulatory monitoring.
Article
Full-text available
Variable and feature selection have become the focus of much research in areas of application for which datasets with tells or hundreds of thousands of variables are available. These areas include text processing of internet documents, gene expression array analysis, and combinatorial chemistry. The objective of variable selection is three-fold: improving the prediction performance of the predictors, providing faster and more cost-effective predictors, and providing a better understanding of the underlying process that generated the data. The contributions of this special issue cover a wide range of aspects of such problems: providing a better definition of the objective function, feature construction, feature ranking, multivariate feature selection, efficient search methods, and feature validity assessment methods.
Article
Full-text available
Objectives: To determine if there are changes in temporal gait parameters with a focus on pelvis when comparing overground and treadmill ambulation, and to assess the effect of sex. Design: An observational study employing motion analysis techniques to evaluate pelvic movement during gait. Setting: University biomechanics laboratory. Participants: Men (n=8; 22.5±3.0 y) and women (n=6; 23.8±4.1 y). Interventions: Not applicable. Main outcome measures: Cadence, stride time, stance phase percent, and pelvic tilt, obliquity, and rotation parameters throughout the gait cycle were assessed during overground and treadmill walking. Kinematic data were recorded using a passive full body marker based motion analysis system. While an independent sample t test was used to determine if differences in walking speed were evident between sexes, a 2-way, repeated-measures analysis of variance was performed to examine the effect of walking mode and sex on each dependent variable. Results: Significant differences (P<.05) between overground and treadmill walking for the temporal parameters analyzed were evident for both sexes. A lower pelvic obliquity motion for treadmill walking when compared with overground walking was evident for both sexes, and the pelvic rotation movement pattern showed the greatest difference between walking modes. The majority of the significant differences between sexes were of a magnitude greater than the differences between overground and treadmill walking. Conclusions: The differences in temporal and angular kinematics identified in the present study should be considered when treadmills are used in a rehabilitation program.
Article
The design of pattern recognition-based myoelectric interfaces has been heavily explored and contested in the research literature. A considerable proportion of the performance of these interfaces has been linked to the quality of the feature extraction (FE) stage used to describe the underlying myoelectric signal. In this paper, we address two important factors of FE that have not been fully exploited; 1) Spatial focus - traditional FE methods focus mainly on the concatenation of features extracted from individual channels and 2) Temporal focus - available FE methods are cross-sectional in nature, largely ignoring the temporal information that may exist between feature windows. To overcome these limitations, several spatiotemporal FE methods have been proposed including hand-crafted and deep learning (DL) models, with the latter showing significant performance enhancements at the cost of increased computational burden. This paper tackles the aforementioned limitations by 1) proposing novel extensions to simple time-domain FE methods, including the waveform length, zero crossings and root mean square, that can capture the relation between any number of channels, and 2) leveraging the long short-term memory concepts from deep neural networks to build a recursive framework that overcomes the cross-sectional nature of traditional methods. The advantages offered by the proposed Recursive Multi-Signal Temporal Fusion (RMTF) features include improved performance, competing with state-of-the-art FE methods without the computational costs associated with leading DL models, and the simplicity of the concepts making them suitable for real-time implementations. Experiments on 65 intact-limbed and amputee subjects reveal an approximate average of 15% reduction in classification errors as compared to models built with other feature sets.
Article
Using multiple smart devices, such as smartphone and smartwatch simultaneously, is becoming a popular life style with the popularity of wearables. This multiple-sensor setting provides new opportunities for enhanced user trait analysis via multiple data fusion. In this study, we explore the task of gender recognition by using motion data collected from multiple smart devices. Specifically, motion data are collected from smartphone and smart band simultaneously. Motion features are extracted from the collected motion data according to three aspects: time, frequency, and wavelet domains. We present a feature selection method considering the redundancies between motion features. Gender recognition is performed using four supervised learning methods. Experimental results demonstrate that using motion data collected from multiple smart devices can significantly improve the accuracy of gender recognition. Evaluation of our method on a dataset of 56 subjects shows that it can reach an accuracy of 98.7% compared with the accuracies of 93.7% and 88.2% when using smartphone and smart band individually.
Article
This paper presents a gait sub-phase detection and prediction approach using surface electromyogram (sEMG) signals, pressure sensors, and the knee angle for a lower-limb power-assist robot. Pattern recognition and machine learning models using sEMG signals have several inherent problems for gait sub-phase detection. These problems are due to recognition delay, lack of consideration for the unique characteristics of sEMG signals based on the subject, and meaningless features. To solve these problems, we propose a new labeling technique based on the heel and toe, a muscle and feature selection, a user-adaptive classifier using a weighted voting technique to achieve gait sub-phase detection, and a gait sub-phase prediction technique using interpolation. Experimental results show that the average accuracies of the proposed labeling, the muscle and feature selection, and the user-adaptive classifier using weighted voting are 7%, 12%, and 17% better, respectively, than the existing methods using physical sensors. Results also show that the average prediction time of the proposed method is 80% faster than the existing methods.
Conference Paper
Robotic systems for rehabilitation of movement disorders and motion assistance are gaining increased attention. Robust classification of motion data as well as reliable recognition of the user’s intended movement play a major role in order to maximize wearability and effectiveness of such systems. Biological signals like electromyography (EMG) provide a direct connection to the motion intention of the wearer. This paper addresses the classification of stance phase and swing phase during healthy human gait based on the muscle activity in both legs using the theory of Support Vector Machines (SVM). A novel EMG feature calculated from the bilateral EMG signals of muscle pairs is introduced. The presented method shows promising results with classification accuracies of up to 96%.
Conference Paper
ECG signals have been widely studied for knowing heart behavior and following cardiac abnormalities. Last years have emerged new applications where ECG has being used in cryptography and biometrics. The purpose in this paper center around perform two independent experiments taking advantage of the ECG properties. The first experiment is about person authentication and the second experiment covers gender recognition. Both tests are performed extracting the same features and evaluating the classification accuracy with several machine learning algorithms sensing the ECG signals in different body positions. ECG signal contains properties like liveness detection, ubiquity, diffculty of being copied, continuity, and reclaims the mandatory user presence. These properties makes ECG study having the potential of being embedded for smartphone applications in the Internet of Things era. The best accuracy score is over the 98% for ECG authentication and 94% for gender recognition; as the best of our knowledge there is no ECG gender recognition with the algorithms studied in this paper.
Article
Gender classification in smartphones has a lot of potential applications. Specifically, the gender information can be used by expert and intelligent systems that are part of healthcare, smart spaces and biometric-based access control applications. For example, operations of intelligent systems in a smart space can be customized based on gender information to provide an enhanced user experience. Similarly, a biometric system can use gender as a soft biometric trait to improve its user authentication performance. This paper presents an approach for gender classification using users’ gait information captured using the built-in sensors of a smartphone. Histogram of gradient (HG) method is proposed to extract features from the gait data, which includes a set of signals collected from accelerometer and gyroscope sensors of a smartphone. The bootstrap aggregating classifier utilizes the discriminatory information in these features for classification of the gender. The performance of the proposed approach has been evaluated on datasets collected using two different smartphones. These datasets contain a total of 654 gait data from 109 subjects. Our experimental results show that the classification accuracy of the proposed approach is higher than that of the existing methods. Additional experiments performed to examine the effect of variations in walking speed indicate that these variations have a minimal impact on the performance of proposed approach. Furthermore, results from our experiments performed on the gait data collected using two different smartphones suggest that the performance of the proposed algorithm for gender recognition is consistent across the two datasets, achieving classification accuracies of 91.78%, 94.44% and 88.89% on the first dataset and 90.48%, 91.07% and 88.46% on the second dataset for normal, fast and slow walking speeds, respectively. The results of this study are significant as they indicate that gait information captured by the smartphones’ built-in sensors can be used to derive gender information reliably and unobtrusively.
Article
To provide consumers with high-quality and safe tea, an accurate, fast and effective method of discriminating moldy tea based on hyperspectral technology was put forward. Moldy tea with three different degrees was studied in this paper. Hyperspectral images of all samples were firstly acquired by the hyperspectral imaging system, and then the spectral data was extracted from the images. Savitzky-Golay and SNV-Detrending were respectively used to pretreat the spectral data. After that, the random forest-recursive feature elimination and the principal component analysis were adopted for feature selection and feature extraction. By combining different pretreatment and feature screening methods, the softmax model optimised by the gradient descent algorithm was established to identify the tea with different moldy degrees. Comparing the classification accuracy and cost function value of all models, the softmax model based on Savitzky-Golay and SNV-Detrending preprocessing and random forest-recursive feature elimination feature selection performed best, which achieved the identification accuracy of 100% for the training set and 98.5% for the test set. Therefore, it is feasible to identify moldy tea with different degrees by using RF-RFE-softmax model and hyperspectral technology.
Article
This study presents a gait subphase recognition method using an electromyogram (EMG) with a signal graph matching (ESGM) algorithm. Existing pattern recognition and machine learning using EMG signals has several innate problems in gait subphase detection. With respect to time domain features, their feature values may be analogous because two different gait steps may have similar muscle activation. In addition, the current gait subphase might not be recognized until the next gait subphase passes because the window size needed for feature extraction is larger than the period of the gait subphase. The ESGM algorithm is a new approach that compares reference EMG signals and input EMG signals according to time variance to solve these problems and considers variations of physiological muscle activity. We also determined all the elements of the ESGM algorithm using kinematic gait analysis and optimized the algorithm using experiments. Therefore, the ESGM algorithm reflects better timing characteristics of EMG signals than the time domain feature extraction algorithm. In addition, it can provide real-time and user-adaptive recognition of the gait subphase by using only EMG signals. Experimental results show that the average accuracy of the proposed method is 13% better than existing methods and the average detection latency of the proposed method was 5.5 times lower than existing methods.
Article
This letter presents lower-limb human motion detection using an surface electromyogram (sEMG) with a top and slope (TAS) feature extraction algorithm. Lower-limb human motion detection using sEMG signal is generally divided into gait sub-phase detection, locomotion mode recognition, and mode change detection. Existing feature extraction algorithms using sEMG signal have several innate problems in recognizing lower-limb human motion detection. With respect to time-domain features, their values may be analogous because two different gait sub-phases and locomotion mode may have similar muscle activity pattern. Therefore, it is not easy to select the proper feature set of sEMG signals. The TAS feature extraction algorithm reflects better timing characteristics of sEMG signals than the existing time-domain feature extraction algorithm. Therefore, it can provide high accuracy in lower-limb human motion detection. Experimental results show that the average detection accuracy values of the proposed method in terms of the gait sub-phase detection, locomotion mode recognition and mode change detection are increased by 8%, 5%, and 4% better than those of feature values of the Willison Amplitude, respectively.
Conference Paper
This paper presents an approach for gender recognition using behavioral biometrics in smartphones. Specifically, this work investigates gender recognition using gait data acquired from the inbuilt accelerometer and gyroscope sensors of a smartphone. The proposed approach involves computation of curvature of the gait signals. In order to capture the local variations of estimated curvatures, we employed histogram features of multi-level local pattern (MLP) and local binary pattern (LBP). In this work, support vector machine (SVM) and aggregate bootstrapping (bagging) classifiers are employed for identification of gender based on the extracted features. Performance evaluation of the proposed approach on a database of 252 gait data collected from 42 subjects yielded promising results. Our experimental results also show that MLP performs better than LBP for feature extraction, while bagging outperforms SVM for classification.
Article
Land use classification plays an important role in adjusting land structure and developing land resources reasonably, especially in the farming area. The objective of this research is to choose an appropriate method to classify land use type in the farming area. A new classification method, random forest (RF) classifier, was applied to make land use mapping in agricultural cultivation region with multi-source information, including multi-seasonal spectrum, texture and topographic information. The best classification scheme was chosen to extract land use information, and RF algorithm was used to reduce the dimension of characteristics variables. The RF algorithm, support vector machine, and maximum likelihood classification were used to map agricultural land use, and the applicability of these three different classification methods was analyzed. The result shows that RF classification of land use classification with multi-source information effects best, the overall accuracy and Kappa coefficient are 85.54% and 0.8359 respectively. Feature selection method from RF algorithm can effectively reduce the data dimension and ensure the accuracy of classification at the same time. Compared with these three classification methods, RF algorithm performs the highest overall accuracy of 81.08%, which is respectively 9.46% and 5.27% higher than support vector machine and maximum likelihood classification. It is an effective scheme that makes land use classification in the farming area using RF classifier with multi-source information. It provides a fast and feasible method for the division of land use types. © 2016, Chinese Society of Agricultural Machinery. All right reserved.
Article
Deep learning methods are able to automatically discover better representations of the data to improve the performance of the classifiers. However, in computer vision tasks, such as the gender recognition problem, sometimes it is difficult to directly learn from the entire image. In this work we propose a new model called Local Deep Neural Network (Local-DNN), which is based on two key concepts: local features and deep architectures. The model learns from small overlapping regions in the visual field using discriminative feed-forward networks with several layers. We evaluate our approach on two well-known gender benchmarks, showing that our Local-DNN outperforms other deep learning methods also evaluated and obtains state-of-the-art results in both benchmarks.
Article
The present study was designed to achieve a comprehensive analysis of gender-related differences in the myoelectric activity of lower limb muscles during normal walking at self-selected speed and cadence, in terms of muscle activation patterns and occurrence frequencies. To this aim, statistical gait analysis (SGA) of surface EMG signal from tibialis anterior (TA), gastrocnemius lateralis (GL), rectus femoris (RF), biceps femoris (BF) and vastus lateralis (VL) was performed in 11 female (F-group) and 11 male (M-group) age-matched healthy young adults. SGA is a recent methodology performing a statistical characterization of gait, by averaging spatio-temporal and sEMG-based parameters over numerous strides. Findings showed that males and females walk at the same comfortable speed, despite the significantly lower height and higher cadence detected in females. No significant differences in muscle onset/offset were detected between groups. The analysis of occurrence frequencies of muscle activity showed no significant differences in BF and RF, between groups. Conversely, in F-group, compared with M-group, GL, TA and VL showed a significantly higher occurrence frequency in the modalities with a high number of activations, and a significantly lower occurrence frequency in the modalities with a low number of activations. These findings indicate a propensity of females for a more complex recruitment of TA, GL and VL during walking, compared to males. The observed differences recommend the suitability of developing electromyographic databases, separated for males and females.
Article
In this paper, a novel method is proposed to build an ensemble of classifiers by using a feature selection schema. The feature selection schema identifies the best feature sets that affect the arrhythmia classification. Firstly, a number of feature subsets are extracted by applying the feature selection schema to the original dataset. Then classification models are built by using the each feature subset. Finally, we combine the classification models by adopting a voting approach to form a classification ensemble. The voting approach in our method involves both classification error rate and feature selection rate to calculate the score of the each classifier in the ensemble. In our method, the feature selection rate depends on the extracting order of the feature subsets. In the experiment, we applied our method to arrhythmia dataset and generated three top disjointed feature sets. We then built three classifiers based on the top-three feature subsets and formed the classifier ensemble by using the voting approach. Our method can improve the classification accuracy in high dimensional dataset. The performance of each classifier and the performance of their ensemble were higher than the performance of the classifier that was based on whole feature space of the dataset. The classification performance was improved and a more stable classification model could be constructed with the proposed approach.
Conference Paper
Recently, there have been many studies to bionic leg to help rehabilitation of lower limb amputees. These powered artificial prosthesis uses physical sensors and repeats trained gait movements in correction process. Therefore, existing methods have to use equal gait speed and movement regardless of person's intention. To solve this problem, in this paper, we propose detailed gait phase recognition method to classify three stance and one swing sub-phases by using heel and toe classifiers and classification matrix. EMG signals are extracted from four body locations of thigh such as Recus femoris, Vastus lateralis, Vastus medialis, Semitendinosus. And then we calculate feature values of the time-domain(MAV, VAR, WL, RMS, SSI) and apply two step classifiers. Experimental result shows that the accuracy of SVM heel classifier is 88%, that of SVM toe classifier is 94% when supervised extracted samples are used, and the accuracy of SVM heel classifier is 78.7%, that of SVM toe classifier is 79.3% when sequentially extracted samples are used, and the average accuracy of the proposed method(SVM) is 79% while that of existing method(SVM) is 61% in case of 4 sub-phase classification when sequentially extracted samples are used.
Article
Various types of data sources have been used to recognize user intent for volitional control of powered artificial legs. However, there is still a debate on what exact data sources are necessary for accurately and responsively recognizing the user's intended tasks. Motivated by this widely interested question, in this study we aimed to 1) investigate the usefulness of different data sources commonly suggested for user intent recognition and 2) determine an informative set of data sources for volitional control of prosthetic legs. The studied data sources included eight surface electromyography (EMG) signals from the residual thigh muscles of transfemoral (TF) amputees, ground reaction forces/moments from a prosthetic pylon, and kinematic measurements from the residual thigh and prosthetic knee. We then ranked and included data sources based on the usefulness for user intent recognition and selected a reduced number of data sources that ensured accurate recognition of the user's intended task by using three source selection algorithms. The results showed that EMG signals and ground reaction forces/moments were more informative than prosthesis kinematics. Nine to eleven of all the initial data sources were sufficient to maintain 95% accuracy for recognizing the studied seven tasks without missing additional task transitions in real time. The selected data sources produced consistent system performance across two experimental days for four recruited TF amputee subjects, indicating the potential robustness of the selected data sources. Finally, based on the study results, we suggested a protocol for determining the informative data sources and sensor configurations for future development of volitional control of powered artificial legs.
Article
Feature extraction is a significant method to extract the useful information which is hidden in surface electromyography (EMG) signal and to remove the unwanted part and interferences. To be successful in classification of the EMG signal, selection of a feature vector ought to be carefully considered. However, numerous studies of the EMG signal classification have used a feature set that have contained a number of redundant features. In this study, most complete and up-to-date thirty-seven time domain and frequency domain features have been proposed to be studied their properties. The results, which were verified by scatter plot of features, statistical analysis and classifier, indicated that most time domain features are superfluity and redundancy. They can be grouped according to mathematical property and information into four main types: energy and complexity, frequency, prediction model, and time-dependence. On the other hand, all frequency domain features are calculated based on statistical parameters of EMG power spectral density. Its performance in class separability viewpoint is not suitable for EMG recognition system. Recommendation of features to avoid the usage of redundant features for classifier in EMG signal classification applications is also proposed in this study.
Article
The clinical application of robotic technology to powered prosthetic knees and ankles is limited by the lack of a robust control strategy. We found that the use of electromyographic (EMG) signals from natively innervated and surgically reinnervated residual thigh muscles in a patient who had undergone knee amputation improved control of a robotic leg prosthesis. EMG signals were decoded with a pattern-recognition algorithm and combined with data from sensors on the prosthesis to interpret the patient's intended movements. This provided robust and intuitive control of ambulation--with seamless transitions between walking on level ground, stairs, and ramps--and of the ability to reposition the leg while the patient was seated.
Article
It is important to analyze the characteristics of normal gait in clinical and biomechanical aspects. Although gait characteristics can be varied by anthropometric, racial and cultural factors, gait studies have primarily been undertaken in Western countries. The present study conducted a gait analysis for Korean people and compared the gait characteristics with those of Western people. A total of 32 Koreans subjects (20 males and 12 females) participated in the gait experiment and their spatio-temporal and kinematic/kinetic characteristics were analyzed. The comparison of the gait characteristics between Korean and Western people revealed that the stride length and walking speed of Korean subjects were significantly lower than those observed in Western studies by 7–25% and 14–42%, respectively. The knee abduction moment of the Korean subjects was larger than those of Western people, while the other moments (hip moment in the sagittal and frontal plane, knee and ankle moment in the sagittal plane) were smaller than those of Western people. There were also differences in ranges of motion between gait studies; however, most motion patterns and excursions were similar.
Conference Paper
The application of hidden Markov model (HMM) to recognize gait phase using electromyographic (EMG) signals is described. Four time-domain features are extracted within a time segment of each channel of EMG signals to preserve pattern structure. According to the division of the gait cycle, the structure of HMM is determined, in which each state is associated with a gait phase. A modified Baum-Welch algorithm is used to estimate the parameter of HMM. And Viterbi algorithm achieves the phase recognition by finding the best state sequence to assign corresponding phases to the given segments. The feature set and data segmentation manner yielded high rate of accuracy are ascertained through evaluation experiments.
Article
Background: Women have higher non-contact anterior cruciate ligament injury rate than men do in sport activities. Non-contact anterior cruciate ligament injuries frequently occur in sports requiring cutting tasks. Alternated motor control strategies have identified as a potential risk factor for the non-contact anterior cruciate ligament injuries. The purpose of this study was to compare the patterns of knee kinematics and electromyographic activities in running, side-cutting, and cross-cutting between men and women recreational athletes. Methods: Three-dimensional kinematic data of the knee and electromyographic data of selected muscles across the knee joint were collected for 11 men and 9 women recreational athletes in running, side-cutting, and cross-cutting. Regression analyses with dummy variables for comparison of knee motion patterns between men and women. Results: Women tend to have less knee flexion angles, more knee valgus angles, greater quadriceps activation, and lower hamstring activation in comparison to men during the stance phase of each of the three athletic tasks. Literatures suggest these alternated knee motion patterns of women tend to increase the load on the anterior cruciate ligament. Conclusion: Women on average may have certain motor control strategies that may alter their knee motion patterns. Women's altered knee motion patterns may tend to increase the load on the anterior cruciate ligament in the selected athletic tasks, which may contribute to the increased anterior cruciate ligament injury rate among women. Relevance: Non-contact anterior cruciate ligament injuries frequently occur in sports. Altered motor control strategies and lower extremity motion patterns are likely to play an important role in non-contact anterior cruciate ligament injuries. Non-contact anterior cruciate ligament injuries may be prevented by correcting altered motor control strategies and associated lower extremity motion patterns through certain training programs.
Article
We present the results of a study in which we investigated the patterns and ranges of movement of the lower thoracic and lumbar spinal segments and the pelvis in subjects walking at two self-selected speeds. Our subjects were aged from 20 to 82 years and both genders were equally represented. Measurements were carried out using a video-based system. We detected increased range of motion in each segment with increased walking speed, few gender-related differences in patterns or ranges of motion and significant reduction in spinal range of motion with advancing age. Our findings suggest, however, that these age-related changes are more likely to be step-length dependent than an intrinsic feature of aging.
Conference Paper
An inertial navigation system for pedestrian position tracking is proposed, where the position is computed using inertial and magnetic sensors on shoes. Using the fact that there is a zero velocity interval in each stride, estimation errors are reduced. When implementing this zero velocity updating algorithm, it is important to know when is the zero velocity interval. The gait states are modeled as a Markov process and gait state is estimated using the hidden Markov model filter. With this gait estimation, the zero velocity interval is more accurately estimated, which helps to reduce the position estimation error.
Article
DNA micro-arrays now permit scientists to screen thousands of genes simultaneously and determine whether those genes are active, hyperactive or silent in normal or cancerous tissue. Because these new micro-array devices generate bewildering amounts of raw data, new analytical methods must be developed to sort out whether cancer tissues have distinctive signatures of gene expression over normal tissues or other types of cancer tissues. In this paper, we address the problem of selection of a small subset of genes from broad patterns of gene expression data, recorded on DNA micro-arrays. Using available training examples from cancer and normal patients, we build a classifier suitable for genetic diagnosis, as well as drug discovery. Previous attempts to address this problem select genes with correlation techniques. We propose a new method of gene selection utilizing Support Vector Machine methods based on Recursive Feature Elimination (RFE). We demonstrate experimentally that the genes selected by our techniques yield better classification performance and are biologically relevant to cancer. In contrast with the baseline method, our method eliminates gene redundancy automatically and yields better and more compact gene subsets. In patients with leukemia our method discovered 2 genes that yield zero leave-one-out error, while 64 genes are necessary for the baseline method to get the best result (one leave-one-out error). In the colon cancer database, using only 4 genes our method is 98% accurate, while the baseline method is only 86% accurate.