Michela Milano

• PhD
• University of Bologna

Traditional university lessons do not provide students with the opportunity to put theoretical concepts into practice. Project-Based Learning is designed to involve students through the proposition of real-word problems in the form of a project. The main objective of this work is to improve several aspects of the student's learning experience (e.g....

This paper focuses on the potential of Generative Artificial Intelligence (AI), particularly Large Language Models (LLMs), in the still unexplored domain of robotic dance creation. In particular, we assess whether a LLM (GPT-3.5 turbo) can create robotic dance choreographies, and we investigate if the feedback provided by human creators can improve...

Recent advancements in Artificial Intelligence in Education (AIEd) have revolutionized educational practices using machine learning to extract insights from students' activities and behaviours. Performance prediction, a key domain within AIEd, aims to enhance student achievement levels and address sustainable development goals related to education,...

Symbolic Artificial Intelligence (AI) techniques in robotic choreography contribute to the broader field of human-robot interaction and address the main limitations of learning approaches in terms of computational requirements. A general symbolic approach leveraging dance sequence rules and constraints can (1) reduce computational effort and (2) al...

Artificial Intelligence (AI) has gradually attracted attention in the field of artistic creation, resulting in a debate on the evaluation of AI artistic outputs. However, there is a lack of common criteria for objective artistic evaluation both of human and AI creations. This is a frequent issue in the field of dance, where different performance me...

A significant challenge in detecting and mitigating bias is creating a mindset amongst AI developers to address unfairness. The current literature on fairness is broad, and the learning curve to distinguish where to use existing metrics and techniques for bias detection or mitigation is difficult. This survey systematises the state-of-the-art about...

The interplay between Machine Learning (ML) and Constrained Optimization (CO) has recently been the subject of increasing interest, leading to a new and prolific research area covering (e.g.) Decision Focused Learning and Constrained Reinforcement Learning. Such approaches strive to tackle complex decision problems under uncertainty over multiple s...

In recent years, the uptake of Artificial Intelligence (AI) in industry is increasing. For many AI techniques, like Deep Learning, optimization, planning, etc., computational and storage requirements are significant. The problem of determining what is the right hardware (HW on premise or on the cloud) architecture and its dimensioning for AI algori...

A typical Virtual Power Plant (VPP) has a distributed architecture, composed by a central control system and decentralized control units, which coordinates and aggregates local resources. A key aspect of these distributed energy systems is the flexibility offered to the market. This flexibility is considered as the difference between the (partially...

Recent research has shown how Deep Neural Networks trained on historical solution pools can tackle CSPs to some degree, with potential applications in problems with implicit soft and hard constraints. In this paper, we consider a setup where one has offline access to symbolic, incomplete, problem knowledge, which cannot however be employed at searc...

In their quest toward Exascale, High Performance Computing (HPC) systems are rapidly becoming larger and more complex, together with the issues concerning their maintenance. Luckily, many current HPC systems are endowed with data monitoring infrastructures that characterize the system state, and whose data can be used to train Deep Learning (DL) an...

Adding constraint support in Machine Learning has the potential to address outstanding issues in data-driven AI systems, such as safety and fairness. Existing approaches typically apply constrained optimization techniques to ML training, enforce constraint satisfaction by adjusting the model design, or use constraints to correct the output. Here, w...

Regularization-based approaches for injecting constraints in Machine Learning (ML) were introduced (see e.g. [2, 5, 8, 9]) to improve a predictive model via domain knowledge. Given the recent interest in ethical and trustworthy AI, however, several works are resorting to these approaches for enforcing desired properties over a ML model (e.g. fairne...

The advent of the coronavirus pandemic has sparked the interest in predictive models capable of forecasting virus-spreading, especially for boosting and supporting decision-making processes. In this paper, we will outline the main Deep Learning approaches aimed at predicting the spreading of a disease in space and time. The aim is to show the emerg...

This paper considers multi-stage optimization problems under uncertainty that involve distinct offline and online phases. In particular it addresses the issue of integrating these phases to show how the two are often interrelated in real-world applications. Our methods are applicable under two (fairly general) conditions: 1) the uncertainty is exog...

Demand Side Management (DSM) is usually considered as a process of energy consumption shifting from peak hours to off-peak times. DSM does not always reduce total energy consumption, but it helps to meet energy demand and supply. For example, it balances variable generation from renewables (such as solar and wind) when energy demand differs from re...

In order to find decisions in finance, regulations, politics including long term strategies for the electricity system and strategic planning on the industrial side, a holistic view on the overall energy system and markets is required at different levels of detail.

Optimization problems under uncertainty are traditionally solved either via offline or online methods. Offline approaches can obtain high-quality robust solutions, but have a considerable computational cost. Online algorithms can react to unexpected events once they are observed, but often run under strict time constraints, preventing the computati...

Regularization-based approaches for injecting constraints in Machine Learning (ML) were introduced to improve a predictive model via expert knowledge. We tackle the issue of finding the right balance between the loss (the accuracy of the learner) and the regularization term (the degree of constraint satisfaction). The key results of this paper is t...

Deep Learning (DL) models proved themselves to perform extremely well on a wide variety of learning tasks, as they can learn useful patterns from large data sets. However, purely data-driven models might struggle when very difficult functions need to be learned or when there is not enough available training data. Fortunately, in many domains prior...

Given enough data, Deep Neural Networks (DNNs) are capable of learning complex input-output relations with high accuracy. In several domains, however, data is scarce or expensive to retrieve, while a substantial amount of expert knowledge is available. It seems reasonable that if we can inject this additional information in the DNN, we could ease t...

Methods for taking into account external knowledge in Machine Learning models have the potential to address outstanding issues in data-driven AI methods, such as improving safety and fairness, and can simplify training in the presence of scarce data. We propose a simple, but effective, method for injecting constraints at training time in supervised...

Machine Learning (ML) models are very effective in many learning tasks, due to the capability to extract meaningful information from large data sets. Nevertheless, there are learning problems that cannot be easily solved relying on pure data, e.g. scarce data or very complex functions to be approximated. Fortunately, in many contexts domain knowled...

The growing demands of the worldwide IT infrastructure stress the need for reduced power consumption, which is addressed in so-called transprecision computing by improving energy efficiency at the expense of precision. For example, reducing the number of bits for some floating-point operations leads to higher efficiency, but also to a non-linear de...

Machine Learning (ML) models are very effective in many learning tasks, due to the capability to extract meaningful information from large data sets. Nevertheless, there are learning problems that cannot be easily solved relying on pure data, e.g. scarce data or very complex functions to be approximated. Fortunately, in many contexts domain knowled...

Power consumption is an increasingly limiting factor in modern ICT infrastructure, especially in the context of High Performance Computing. Common strategies to curb energy consumption are power capping, i.e. constraining the system power consumption within certain power budget, and Dynamic Voltage/Frequency Scaling, i.e. reducing the computing ele...

High Performance Computing (HPC) systems are complex machines with heterogeneous components that can break or malfunction. Automated anomaly detection in these systems is a challenging and critical task, as HPC systems are expected to work 24/7. The majority of the current state-of-the-art methods dealing with this problem are Machine Learning tech...

When optimising under uncertainty, it is desirable that solutions are robust to unexpected disruptions and changes. A possible formalisation of robustness is given by super solutions. An assignment to a set of decision variables is an (a, b, c) super solution if any change involving at most a variables can be repaired by changing at most b other va...

Sampling-based anticipatory algorithms can be very effective at solving online optimization problems under uncertainty, but their computational cost may be prohibitive in some cases. Given an arbitrary anticipatory algorithm, we present three methods that allow to retain its solution quality at a fraction of the online computational cost, via a sub...

Anomaly detection in supercomputers is a very difficult problem due to the big scale of the systems and the high number of components. The current state of the art for automated anomaly detection employs Machine Learning methods or statistical regression models in a supervised fashion, meaning that the detection tool is trained to distinguish among...

Designing and evaluating energy policies is a difficult challenge because the energy sector is a complex system that cannot be adequately understood without using models merging economic, social and individual perspectives. Appropriate models allow policy makers to assess the impact of policy measures, satisfy strategic objectives and develop susta...

Energy efficiency is of paramount importance for the sustainability of high performance computing (HPC) systems. Energy consumption limits the peak performance of supercomputers and accounts for a large share of Total Cost of Ownership (TCO). Consequently, system owners and final users have started exploring mechanisms to trade off performance for...

Anomaly detection in supercomputers is a very difficult problem due to the big scale of the systems and the high number of components. The current state of the art for automated anomaly detection employs Machine Learning methods or statistical regression models in a supervised fashion, meaning that the detection tool is trained to distinguish among...

In this work we present two general techniques to deal with multi-stage optimization problems under uncertainty, featuring off-line and on-line decisions. The methods are applicable when: 1) the uncertainty is exogenous; 2) there exists a heuristic for the on-line phase that can be modeled as a parametric convex optimization problem. The first tech...

In the past few years, the area of Machine Learning (ML) has witnessed tremendous advancements, becoming a pervasive technology in a wide range of applications. One area that can significantly benefit from the use of ML is Combinatorial Optimization. The three pillars of constraint satisfaction and optimization problem solving, i.e., modeling, sear...

Deep Neural Networks (DNNs) have been shaking the AI scene, for their ability to excel at Machine Learning tasks without relying on complex, hand-crafted, features. Here, we probe whether a DNN can learn how to construct solutions of a CSP, without any explicit symbolic information about the problem constraints. We train a DNN to extend a feasible...

Energy efficiency is of paramount importance for the sustainability of HPC systems. Energy consumption limits the peak performance of supercomputers and accounts for a large share of total cost of ownership. Consequently, system owners and final users have started exploring mechanisms to trade off performance for power consumption, for example thro...

Optimization problems under uncertainty arise in many application areas and their solution is very challenging. We propose here methods that merge off-line and on-line decision stages: we start with a two stage off-line approach coupled with an on-line heuristic. We improve this baseline in two directions: (1) by replacing the on-line heuristics wi...

Supercomputer installed capacity worldwide increased for many years and further growth is expected in the future. The next goal for high performance computing (HPC) systems is reaching Exascale. The increase in computational power threatens to lead to unacceptable power demands, if future machines will be built using current technology. Therefore r...

This paper is motivated by the concept that the successful, effective, and sustainable implementation of the smart city paradigm requires a close cooperation among researchers with different, complementary interests and, in most cases, a multidisciplinary approach. It first briefly discusses how such a multidisciplinary methodology, transversal to...

Virtual Power Plants (VPP) are one of the main components of future smart electrical grids, connecting and integrating several types of energy sources, loads and storage devices. A typical VPP is a large industrial plant with high (partially shiftable) electric and thermal loads, renewable energy generators and electric and thermal storages. Optimi...

Recently, a number of noteworthy results have been achieved in various fields of artificial intelligence, and many aspects of the problem solving process have received significant attention by the scientific community. In this context, the extraction of comprehensive knowledge suitable for problem solving and reasoning, from textual and pictorial p...

Demand response mechanisms and load control in the electricity market represent an important area of research at the international level: the trend towards competition and market liberalization has led to the development of methodologies and tools to support energy providers. Demand side management helps energy suppliers to reduce the peak demand a...

Scheduling and dispatching are critical enabling technologies in supercomputing and grid computing. In these contexts, scalability is an issue: we have to allocate and schedule up to tens of thousands of tasks on tens of thousands of resources. This problem scale is out of reach for complete and centralized scheduling approaches. We propose a distr...

Many real world cyclic scheduling problems involve applications that need to be repeated with different periodicity. For example, multirate control systems present multiple control loops that are organized hierarchically: the higher-level loop responds to the slower system dynamics and typically its period can be a few orders of magnitude longer th...

Scheduling and dispatching are critical enabling technologies in supercomputing and grid computing. In these contexts, scalability is an issue: we have to allocate and schedule up to tens of thousands of tasks on tens of thousands of resources. This problem scale is out of reach for complete and centralized scheduling approaches. We propose a dist...

Power consumption is a critical aspect for next generation High Performance Computing systems: Supercomputers are expected to reach Exascale in 2023 but this will require a significant improvement in terms of energy efficiency. In this domain, power-capping can significant increase the final energy-efficiency by cutting cooling effort and worst-cas...

Demand Response mechanisms and load control in the electricity market represent an important area of research at international level, and the market liberalization is opening new perspectives. This calls for the development of methodologies and tools that energy providers can use to define specific business models. In this work we develop an optimi...

One of the biggest challenges in the design of real-world decision support systems is coming up with a good combinatorial optimization model. Often enough, accurate predictive models (e.g. simulators) can be devised, but they are too complex or too slow to be employed in combinatorial optimization.In this paper, we propose a methodology called Empi...

Scheduling and dispatching tools for High-Performance Computing (HPC) machines have the key role of mapping jobs to the available resources, trying to maximize performance and Quality-of-Service (QoS). Allocation and Scheduling in the general case are well-known NP-hard problems, forcing commercial schedulers to adopt greedy approaches to improve p...

While most standardized vehicular communication applications aim on increasing traffic safety, the exchange of messages between vehicles and the environment may be used for other purposes at no additional hardware costs as well. One possible area of such applications is traffic management. Traffic management requires data about the state of the roa...

This paper presents a Traffic Lights control system, inspired by Swarm intelligence methodologies, in which every intersection controller makes independent decisions to pursue common goals and is able to improve the global traffic performance. The solution is low cost and widely applicable to different urban scenarios. This work is developed within...

Sustainable energy policies are becoming of paramount importance for our future, shaping the environment around us, underpinning economic growth, and increasingly affecting the geopolitical considerations of governments world-wide. Renewable energy diffusion and energy efficiency measures are key for obtaining a transition toward low carbon energy...

The aim of this document is to indicate how the results of COLOMBO can be exploited. The document does not only focus on the duration of the project, but also assesses the possibilities beyond the end at 31st of October 2015. A good reference for the exploitability of the COLOMBO results is the Technology Readiness Level (TRL), which is used in the...

Power consumption is a key factor in modern ICT infrastruc- ture, especially in the expanding world of High Performance Computing, Cloud Computing and Big Data. Such consumption is bound to become an even greater issue as supercomputers are envisioned to enter the Ex- ascale by 2020, granted that they obtain an order of magnitude energy efficiency...

This paper is about characterizing the expected makespan of a Partial Order Schedule (POS) under duration uncertainty. Our analysis is based on very general assumptions about the uncertainty: in particular, we assume that only the min, max, and average durations are known. This information is compatible with a whole range of values for the expected...

While most standardized vehicular communication applications aim on increasing traffic safety, the exchange of messages between vehicles and the environment may be well used for other purposes at no additional cost as well. One possible area of applications is traffic management. Traffic management requires data about the state of the road network....

Today’s politicians are confronted with new information technologies to tackle complex decision-making problems. In order to make sustainable decisions, a profound analysis of societal problems and possible solutions (policy options) needs to be performed. In this policy-analysis process, different stakeholders are involved. Besides internal direct...

In past papers, we have introduced Empirical Model Learning (EML) as a method to enable Combinatorial Optimization on real world systems that are impervious to classical modeling approaches. The core idea in EML consists in embedding a Machine Learning model in a traditional combinatorial model. So far, the method has been demonstrated by using Neu...

This book constitutes the refereed proceedings of the 7th IFIP WG 8.5 International Conference on Electronic Participation, ePart 2015, held in Thessaloniki, Greece, in August/September 2015. The 12 revised full papers presented were carefully reviewed and selected from 19 submissions. The papers have been organized in the following topical section...

Renewable energy technologies have benefited to varying extent from the support of incentive programmes in-troduced in the industrialised countries over the last 20 years. Understanding the impact of incentive schemas on the adoption of renewable energy sources is a crucial aspect for policy makers. We study in this paper the impact of national inc...

Improving the efficiency of urban vehicular mobility, also via the optimized management of the dynamic behavior of traffic lights with limited infrastructure investments and limited operational costs, is widely recognized as a crucial goal for smart cities, capable of relevant economic impacts in terms of travel time/cost reduction and better susta...

In the era of Cloud Computing, Big Data, and Quantum Physics Simulations, data centers play in the world ICT infrastructure a role as big as (sadly) their power consumption. In many cases, a surpris-ing amount of such consumption is due to idle resources, either intro-duced to face workload peaks or leftovers of workload fragmentation. In this cont...

Policy making is an extremely complex process occurring in changing environments and affecting the three pillars of sustainable development: society, economy and the environment. Each political decision in fact implies some form of social reactions, it affects economic and financial aspects and has substantial environmental impacts. Improving decis...

In chapter 2, a literature research points out a lot of differences regarding to light signals between Austria, Belgium, France, Germany, Italy, Netherlands, and the United Kingdom. Normally the street user can handle the different light signals in a safe way, but when planning light signals certain differences could be recognised. In this chapter,...

A growing number of applications require continuous processing of high-throughput data streams, e.g., financial analysis, network traffic monitoring, or big data analytics. Performing these analyses by using Distributed Stream Processing Systems (DSPSs) in large clusters is emerging as a promising solution to address the scalability challenges pose...

In the policy making process a number of disparate and diverse issues such as economic development, environmental aspects, as well as the social acceptance of the policy, need to be considered. A single person might not have all the required expertises, and decision support systems featuring optimization components can help to assess policies. Leve...

In the context of Scheduling under uncertainty, Partial Order Schedules (POS) provide a convenient way to build flexible solutions. A POS is obtained from a Project Graph by adding precedence constraints so that no resource conflict can arise, for any possible assignment of the activity durations. In this paper, we use a simulation approach to eval...

In this paper, we propose a challenging research direction for Constraint Programming and optimization techniques in general. We address problems where decisions to be taken affect and are affected by complex systems, which exhibit phenomena emerging from a collection of interacting objects, capable to self organize and to adapt their behaviour acc...

This Deliverable 5.1 contains the design and a description of the work undertaken to combine and integrate the engaged software (SUMO, ns3, iCS, PHEM, ...) to suite the investigation tasks of the COLOMBO project. Requirements from the traffic management solutions to be developed in COLOMBO put on the overall simulation system have been collected. T...

Cyclic scheduling problems consist in ordering a set of activities executed indefinitely over time in a periodic fashion, subject to precedence and resource constraints. This class of problems has many applications in manufacturing, embedded systems and compiler design, production and chemical systems. This paper proposes a Constraint Programming a...

Within the COLOMBO project, modern traffic surveillance and traffic light control algorithms based on data obtained from vehicular communications are developed. An evaluation of existing descriptions of traffic light evaluation show that both, common measurement definitions as well as standardised simulation scenarios are missing. We present a defi...

Policy making is a very complex task taking into account several aspects related to sustainability, namely impact on the environments, health of productive sectors, economic implications and social acceptance. Optimization methods could be extremely useful for analysing alternative policy scenarios, but should be complemented with several other tec...

An elegant way to tackle a problem that you cannot solve is to cast it to a problem that you can solve very well. Cyclic Scheduling problems are very simi-lar to Resource Constrained Project Scheduling Prob-lems (RCPSP), except that the project activities are re-peated over time. Due to the similarity, reducing Cyclic Scheduling problems to RCPSPs...

Vorstellung des COLOMBO-Systems zur Reduktion der Emissionen im Straßenverkehr unter Verwendung von Fahrzeug-Fahrzeug- und Fahrzeug-Infrastruktur-Kommunikation bei geringen Ausrüstungsraten.

This is a summary of Lombardi and Milano, 2012, where we propose a novel method for Minimal Critical Set identification, to be used for the solution of scheduling problems via Precedence Constraint Posting. The method is based on a minimum-flow problem and a heuristic minimization step. The proposed approach is much more scalable than enumeration-b...

An elegant way to tackle a problem that you cannot solve is to cast it to a problem that you can solve very well. Cyclic Scheduling problems are very similar to Resource Constrained Project Scheduling Problems (RCPSP), except that the project activities are repeated over time. Due to the similarity, reducing Cyclic Scheduling problems to RCPSPs see...

Designing sustainable energy policies has a strong impact on economy, society and environment. Beside a planning activity, policy makers are called to design a number of implementation instruments to enforce their plans. They encompass subsidies, fiscal incentives, feed in tariffs to name a few. Understanding the impact of these instruments on the...

Effective multicore computing requires to make efficient usage of the computational resources on a chip. Offline mapping and scheduling can be applied to improve the performance, but classical approaches require considerable a priori knowledge of the target application. In a practical setting, precise information is often unavailable; one can then...

Designing sustainable energy policies heavily impacts the economic development, environmental resource management and social acceptance. There are four main steps in the policy making process: planning, environmental assessment, implementation and monitoring. We focus here on the first three steps that are performed ex-ante. We describe in this pap...