Georgios Chalkiadakis - Academia.edu (original) (raw)

Papers by Georgios Chalkiadakis

Lecture Notes in Computer Science, 2022

In this paper we put forward an open multi-agent systems (MAS) architecture for the important and... more In this paper we put forward an open multi-agent systems (MAS) architecture for the important and challenging to engineer vehicleto-grid (V2G) and grid-to-vehicle (G2V) energy transfer problem domains. To promote scalability, our solution is provided in the form of modular microservices that are interconnected using a multi-protocol Internet of Things (IoT) platform. On the one hand, the low-level modularity of Smart Grid services allows the seamless integration of different agent strategies, pricing mechanisms and algorithms; and on the other, the IoT-based implementation offers both direct applicability in realworld settings, as well as advanced analytics capabilities by enabling digital twins models for Smart Grid ecosystems. We describe our MAS/IoTbased architecture and present results from simulations that incorporate large numbers of heterogeneous Smart Grid agents, which might follow different strategies for their decision making tasks. Our framework enables the testing of various schemes in simulation mode, and can also be used as the basis for the implementation of real-world prototypes for the delivery of large-scale V2G/G2V services.

SUMO Conference Proceedings

This paper presents some new developments related to TrafficFluid-Sim, a lane-free microscopic si... more This paper presents some new developments related to TrafficFluid-Sim, a lane-free microscopic simulator that extends the SUMO simulation infrastructure to model lane-free traffic environments, allowing vehicles to be located at any lateral position, disregarding standard notions of car-following and lane-change maneuvers that are typically embedded within a (lanebased) simulator. A dynamic library has been designed for traffic monitoring and lane-free vehicle movement control, one that does not impose any inter-tool “communication” delays that standard practices with the TraCI module introduce; and enables the emulation of vehicleto-vehicle and vehicle-to-infrastructure communication. We first summarize the various core components that constitute our simulator, and then discuss the new capability to utilize the bicycle kinematic model, additionally to the usual double-integrator model, as a more realistic model of vehicle movement dynamics, particularly for a lane-free traffic env...

Proceedings of the 12th Hellenic Conference on Artificial Intelligence

Lecture Notes in Computer Science, 2022

Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence

In this work, we provide novel methods which benefit from obtained probability bounds for assessi... more In this work, we provide novel methods which benefit from obtained probability bounds for assessing the ability of teams of agents to accomplish coalitional tasks. To this end, our first method is based on an improvement of the Paley-Zygmund inequality, while the second and the third ones are devised based on manipulations of the two-sided Chebyshev’s inequality and the Hoeffding’s inequality, respectively. Agents have no knowledge of the amount of resources others possess; and hold private Bayesian beliefs regarding the potential resource investment of every other agent. Our methods allow agents to demand that certain confidence levels are reached, regarding the resource contributions of the various coalitions. In order to tackle real-world scenarios, we allow agents to form overlapping coalitions, so that one can simultaneously be part of a number of coalitions. We thus present a protocol for iterated overlapping coalition formation (OCF), through which agents can complete tasks t...

2021 IEEE International Conference on Imaging Systems and Techniques (IST), 2021

The classification of recyclable materials, and in particular the recovery of plastic, plays an i... more The classification of recyclable materials, and in particular the recovery of plastic, plays an important role in the economy, but also in environmental sustainability. This study presents a novel image classification model that can be efficiently used to distinguish recyclable materials. Building on recent work in deep learning and waste classification, we introduce the so-called "Dual-branch Multi-output CNN", a custom convolutional neural network composed of two branches aimed to i) classify recyclables and ii) distinguish the type of plastic. The proposed architecture is composed of two classifiers trained on two different datasets, so as to encode complementary attributes of the recyclable materials. In our work, the Densenet121, ResNet50 and VGG16 architectures were used on the Trashnet dataset, along with data augmentation techniques, as well as on the WaDaBa dataset with physical variation techniques. In particular, our approach makes use of the joint utilization of the datasets, allowing the learning of disjoint label combinations. Our experiments confirm its effectiveness in the classification of waste material.

The prevalent assumption in hedonic games is that agents are interested solely on the composition... more The prevalent assumption in hedonic games is that agents are interested solely on the composition of their own coalition. Moreover, agent preferences are usually assumed to be known with certainty. In our work, agents have hidden preferences over partitions. We first put forward the formal definition of hedonic games in partition function form (PFF-HGs), and extend well-studied classes of hedonic games to this setting. Then we exploit three well-known supervised learning models, linear regression, linear regression with basis function, and feed forward neural networks, in order to (approximately) extract the unknown hedonic preference relations over partitions. We conduct a systematic evaluation to compare the performance of these models on PFF-HGs; and, in the process, we develop an evaluation metric specifically designed for our problem. Our experimental results confirm the effectiveness of our work.

This paper addresses the problem of collaborative multi-agent autonomous driving of connected and... more This paper addresses the problem of collaborative multi-agent autonomous driving of connected and automated vehicles (CAVs) in lane-free highway scenarios. We eliminate the lane-changing task, i.e., CAVs may be located in any arbitrary lateral position within the road boundaries, hence allowing for better utilization of the available road capacity. As a consequence, vehicles operate in a much more complex environment, and the need for the individual CAVs to select actions that are efficient for the group as a whole is highly desired. We formulate this environment as a multiagent collaboration problem represented via a coordination graph, thus decomposing the problem with local utility functions, based on the interactions between vehicles. We produce a tractable and scalable solution by estimating the joint action of all vehicles via the anytime max-plus algorithm, with local utility functions provided by potential fields, designed to promote collision avoidance. Specifically, the fi...

In this work we address the problem of coordinated consumption shifting for electricity prosumers... more In this work we address the problem of coordinated consumption shifting for electricity prosumers. We show that individual optimization with respect to electricity prices does not always lead to minimized costs, thus necessitating a cooperative approach. A prosumer cooperative employs an internal cryptocurrency mechanism for coordinating members decisions and distributing the collectively generated profits. The mechanism generates cryptocoins in a distributed fashion, and awards them to participants according to various criteria, such as contribution impact and accuracy between stated and final shifting actions. In particular, when a scoring rulesbased distribution method is employed, participants are incentivized to be accurate. When tested on a large dataset with real-world production and consumption data, our approach is shown to provide incentives for accurate statements and increased economic profits for the cooperative.

In recent years, estimating the power output of inherently intermittent and potentially distribut... more In recent years, estimating the power output of inherently intermittent and potentially distributed renewable energy sources has become a major scientific and societal concern. In this paper, we provide an algorithmic framework, along with an interactive web-based tool, to enable short-to-middle term forecasts of photovoltaic (PV) systems and wind generators output. Importantly, we propose a generic PV output estimation method, the backbone of which is a solar irradiance approximation model that incorporates free-to-use, readily available meteorological data coming from online weather stations. The model utilizes non-linear approximation components for turning cloud-coverage into radiation forecasts, such as an MLP neural network with one hidden layer. We present a thorough evaluation of the proposed techniques, and show that they can be successfully employed within a broad geographical region (the Mediterranean belt) and come with specific performance guarantees. Crucially, our met...

In recent years, fake news detection has been an emerging research area. In this paper, we put fo... more In recent years, fake news detection has been an emerging research area. In this paper, we put forward a novel statistical approach for the generation of feature vectors to describe a document. Our so-called class label frequency distance (clfd), is shown experimentally to provide an effective way for boosting the performance of machine learning methods. Specifically, our experiments, carried out in the fake news detection domain, verify that efficient traditional machine learning methods that use our vectorization approach, consistently outperform deep learning methods that use word embeddings for small and medium sized datasets, while the results are comparable for large datasets. In addition, we demonstrate that a novel hybrid method that utilizes both a clfd-boosted logistic regression classifier and a deep learning one, clearly outperforms deep learning methods even in large datasets.

A variety of multiagent systems methods has been proposed for forming cooperatives of interconnec... more A variety of multiagent systems methods has been proposed for forming cooperatives of interconnected agents representing electricity producers or consumers in the Smart Grid. One major problem that arises in this domain is assessing participating agents’ uncertainty, and correctly predicting their future behaviour regarding power consumption shifting actions. In this paper we adopt two stochastic filtering techniques, a Gaussian Process Filter and a Histogram Filter, and use these to effectively monitor the trustworthiness of agent statements regarding their final shifting actions. We incorporate these within a directly applicable scheme for providing electricity demand management services. Experiments were conducted on real-world consumption datasets from Kissamos, a municipality of Crete. Our results confirm that these techniques provide tangible benefits regarding enhanced consumption reduction performance, and increased financial gains for the cooperative.

A variety of multiagent systems methods has been proposed for forming cooperatives of interconnec... more A variety of multiagent systems methods has been proposed for forming cooperatives of interconnected agents representing electricity producers or consumers in the Smart Grid. One major problem that arises in this domain is assessing participating agents uncertainty, and correctly predicting their future behaviour. In this paper, we adopt two stochastic filtering techniques —the Unscented Kalman Filter equipped with Gaussian Processes, and the Histogram Filter— and use these to effectively monitor the trustworthiness of agent statements regarding their final actions. The methods are incorporated within a directly applicable scheme for providing electricity demand management services. Simulation results confirm that these techniques provide tangible benefits regarding enhanced consumption reduction performance, and increased financial gains.

Ο σκοπός αυτής της εργασίας είναι να εμβαθύνει την κατανόησή μας για την παρακμή του Μινωϊκού πολ... more Ο σκοπός αυτής της εργασίας είναι να εμβαθύνει την κατανόησή μας για την παρακμή του Μινωϊκού πολιτισμού, χρησιμοποιώντας ένα υπάρχον πολυπρακτορικό μοντέλο προσομοίωσης (ABM) για να μελετήσει σε ποιο βαθμό η κατακλυσμική έκρηξη του ηφαιστείου της Θήρας επηρέασε την κοινωνική εξέλιξη του Μινωϊκού πολιτισμού. Λαμβάνοντας υπόψη τη γεωργία ως κύρια παραγωγική δραστηριότητα για την διατήρηση του ανθρώπινου πληθυσμού, αξιολογούμε τις επιπτώσεις της έκρηξης του ηφαιστείου πάνω σε διαφορετικά μοντέλα κοινωνικής οργάνωσης, εστιάζοντας στην ευρύτερη περιοχή των Μαλίων της Κρήτης. Τα παραδείγματα κοινωνικής οργάνωσης που εξετάστηκαν είναι εμπνευσμένα από ένα μοντέλο αυτο-οργάνωσης μιας κοινότητας πρακτόρων καθώς και ιδέες από την εξελικτική θεωρία παιγνίων. Οι επιλογές παραμέτρων του πολυπρακτορικού μοντέλου βασίζονται σε αρχαιολογικές θεωρίες και ευρήματα, αλλά δεν είναι μεροληπτικές προς οποιαδήποτε συγκεκριμένη παραδοχή. Αποτελέσματα από διαφορετικά σενάρια προσομοίωσης επιδεικνύουν μια εν...

Tackling the decision-making problem faced by a prosumer (i.e., a producer that is simultaneously... more Tackling the decision-making problem faced by a prosumer (i.e., a producer that is simultaneously a consumer) when selling and buying energy in the emerging smart electricity grid, is of utmost importance for the economic profitability of such a business entity. In this paper, we model, for the first time, this problem as a factored Markov Decision Process. By so doing, we are able to represent the problem compactly, and provide an exact optimal solution via dynamic programming - notwithstanding its large size. Our model successfully captures the main aspects of the business decisions of a prosumer corresponding to a community microgrid of any size. Moreover, it includes appropriate sub-models for prosumer production and consumption prediction. Experimental simulations verify the effectiveness of our approach; and show that our exact value iteration solution matches that of a state-of-the-art method for stochastic planning in very large environments, while outperforming it in terms ...

We propose a novel technique for Active Malware Analysis (AMA) formalized as a Bayesian game betw... more We propose a novel technique for Active Malware Analysis (AMA) formalized as a Bayesian game between an analyzer agent and a malware agent, focusing on the decision making strategy for the analyzer. In our model, the analyzer performs an action on the system to trigger the malware into showing a malicious behavior, i.e., by activating its payload. The formalization is built upon the link between malware families and the notion of types in Bayesian games. A key point is the design of the utility function, which reflects the amount of uncertainty on the type of the adversary after the execution of an analyzer action. This allows us to devise an algorithm to play the game with the aim of minimizing the entropy of the analyzer’s belief at every stage of the game in a myopic fashion. Empirical evaluation indicates that our approach results in a significant improvement both in terms of learning speed and classification score when compared to other state-of-the-art AMA techniques. ACM Refe...

Simulation Modelling Practice and Theory, 2021

Proceedings of the 10th Hellenic Conference on Artificial Intelligence, 2018

In this article, we explore the computation of joint policies for autonomous agents to resolve co... more In this article, we explore the computation of joint policies for autonomous agents to resolve congestions problems in the air traffic management (ATM) domain. Agents, representing flights, have limited information about others' payoffs and preferences, and need to coordinate to achieve their tasks while adhering to operational constraints. We formalize the problem as a multiagent Markov decision process (MDP) towards deciding flight delays to resolve demand and capacity balance (DCB) problems in ATM. To this end, we present multiagent reinforcement learning methods that allow agents to interact and form own policies in coordination with others. Experimental study on real-world cases, confirms the effectiveness of our approach in resolving the demand-capacity balance problem.

Advances in Building Energy Research, 2020

Lecture Notes in Computer Science, 2022

In this paper we put forward an open multi-agent systems (MAS) architecture for the important and... more In this paper we put forward an open multi-agent systems (MAS) architecture for the important and challenging to engineer vehicleto-grid (V2G) and grid-to-vehicle (G2V) energy transfer problem domains. To promote scalability, our solution is provided in the form of modular microservices that are interconnected using a multi-protocol Internet of Things (IoT) platform. On the one hand, the low-level modularity of Smart Grid services allows the seamless integration of different agent strategies, pricing mechanisms and algorithms; and on the other, the IoT-based implementation offers both direct applicability in realworld settings, as well as advanced analytics capabilities by enabling digital twins models for Smart Grid ecosystems. We describe our MAS/IoTbased architecture and present results from simulations that incorporate large numbers of heterogeneous Smart Grid agents, which might follow different strategies for their decision making tasks. Our framework enables the testing of various schemes in simulation mode, and can also be used as the basis for the implementation of real-world prototypes for the delivery of large-scale V2G/G2V services.

SUMO Conference Proceedings

This paper presents some new developments related to TrafficFluid-Sim, a lane-free microscopic si... more This paper presents some new developments related to TrafficFluid-Sim, a lane-free microscopic simulator that extends the SUMO simulation infrastructure to model lane-free traffic environments, allowing vehicles to be located at any lateral position, disregarding standard notions of car-following and lane-change maneuvers that are typically embedded within a (lanebased) simulator. A dynamic library has been designed for traffic monitoring and lane-free vehicle movement control, one that does not impose any inter-tool “communication” delays that standard practices with the TraCI module introduce; and enables the emulation of vehicleto-vehicle and vehicle-to-infrastructure communication. We first summarize the various core components that constitute our simulator, and then discuss the new capability to utilize the bicycle kinematic model, additionally to the usual double-integrator model, as a more realistic model of vehicle movement dynamics, particularly for a lane-free traffic env...

Proceedings of the 12th Hellenic Conference on Artificial Intelligence

Lecture Notes in Computer Science, 2022

Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence

In this work, we provide novel methods which benefit from obtained probability bounds for assessi... more In this work, we provide novel methods which benefit from obtained probability bounds for assessing the ability of teams of agents to accomplish coalitional tasks. To this end, our first method is based on an improvement of the Paley-Zygmund inequality, while the second and the third ones are devised based on manipulations of the two-sided Chebyshev’s inequality and the Hoeffding’s inequality, respectively. Agents have no knowledge of the amount of resources others possess; and hold private Bayesian beliefs regarding the potential resource investment of every other agent. Our methods allow agents to demand that certain confidence levels are reached, regarding the resource contributions of the various coalitions. In order to tackle real-world scenarios, we allow agents to form overlapping coalitions, so that one can simultaneously be part of a number of coalitions. We thus present a protocol for iterated overlapping coalition formation (OCF), through which agents can complete tasks t...

2021 IEEE International Conference on Imaging Systems and Techniques (IST), 2021

The classification of recyclable materials, and in particular the recovery of plastic, plays an i... more The classification of recyclable materials, and in particular the recovery of plastic, plays an important role in the economy, but also in environmental sustainability. This study presents a novel image classification model that can be efficiently used to distinguish recyclable materials. Building on recent work in deep learning and waste classification, we introduce the so-called "Dual-branch Multi-output CNN", a custom convolutional neural network composed of two branches aimed to i) classify recyclables and ii) distinguish the type of plastic. The proposed architecture is composed of two classifiers trained on two different datasets, so as to encode complementary attributes of the recyclable materials. In our work, the Densenet121, ResNet50 and VGG16 architectures were used on the Trashnet dataset, along with data augmentation techniques, as well as on the WaDaBa dataset with physical variation techniques. In particular, our approach makes use of the joint utilization of the datasets, allowing the learning of disjoint label combinations. Our experiments confirm its effectiveness in the classification of waste material.

The prevalent assumption in hedonic games is that agents are interested solely on the composition... more The prevalent assumption in hedonic games is that agents are interested solely on the composition of their own coalition. Moreover, agent preferences are usually assumed to be known with certainty. In our work, agents have hidden preferences over partitions. We first put forward the formal definition of hedonic games in partition function form (PFF-HGs), and extend well-studied classes of hedonic games to this setting. Then we exploit three well-known supervised learning models, linear regression, linear regression with basis function, and feed forward neural networks, in order to (approximately) extract the unknown hedonic preference relations over partitions. We conduct a systematic evaluation to compare the performance of these models on PFF-HGs; and, in the process, we develop an evaluation metric specifically designed for our problem. Our experimental results confirm the effectiveness of our work.

This paper addresses the problem of collaborative multi-agent autonomous driving of connected and... more This paper addresses the problem of collaborative multi-agent autonomous driving of connected and automated vehicles (CAVs) in lane-free highway scenarios. We eliminate the lane-changing task, i.e., CAVs may be located in any arbitrary lateral position within the road boundaries, hence allowing for better utilization of the available road capacity. As a consequence, vehicles operate in a much more complex environment, and the need for the individual CAVs to select actions that are efficient for the group as a whole is highly desired. We formulate this environment as a multiagent collaboration problem represented via a coordination graph, thus decomposing the problem with local utility functions, based on the interactions between vehicles. We produce a tractable and scalable solution by estimating the joint action of all vehicles via the anytime max-plus algorithm, with local utility functions provided by potential fields, designed to promote collision avoidance. Specifically, the fi...

In this work we address the problem of coordinated consumption shifting for electricity prosumers... more In this work we address the problem of coordinated consumption shifting for electricity prosumers. We show that individual optimization with respect to electricity prices does not always lead to minimized costs, thus necessitating a cooperative approach. A prosumer cooperative employs an internal cryptocurrency mechanism for coordinating members decisions and distributing the collectively generated profits. The mechanism generates cryptocoins in a distributed fashion, and awards them to participants according to various criteria, such as contribution impact and accuracy between stated and final shifting actions. In particular, when a scoring rulesbased distribution method is employed, participants are incentivized to be accurate. When tested on a large dataset with real-world production and consumption data, our approach is shown to provide incentives for accurate statements and increased economic profits for the cooperative.

In recent years, estimating the power output of inherently intermittent and potentially distribut... more In recent years, estimating the power output of inherently intermittent and potentially distributed renewable energy sources has become a major scientific and societal concern. In this paper, we provide an algorithmic framework, along with an interactive web-based tool, to enable short-to-middle term forecasts of photovoltaic (PV) systems and wind generators output. Importantly, we propose a generic PV output estimation method, the backbone of which is a solar irradiance approximation model that incorporates free-to-use, readily available meteorological data coming from online weather stations. The model utilizes non-linear approximation components for turning cloud-coverage into radiation forecasts, such as an MLP neural network with one hidden layer. We present a thorough evaluation of the proposed techniques, and show that they can be successfully employed within a broad geographical region (the Mediterranean belt) and come with specific performance guarantees. Crucially, our met...

In recent years, fake news detection has been an emerging research area. In this paper, we put fo... more In recent years, fake news detection has been an emerging research area. In this paper, we put forward a novel statistical approach for the generation of feature vectors to describe a document. Our so-called class label frequency distance (clfd), is shown experimentally to provide an effective way for boosting the performance of machine learning methods. Specifically, our experiments, carried out in the fake news detection domain, verify that efficient traditional machine learning methods that use our vectorization approach, consistently outperform deep learning methods that use word embeddings for small and medium sized datasets, while the results are comparable for large datasets. In addition, we demonstrate that a novel hybrid method that utilizes both a clfd-boosted logistic regression classifier and a deep learning one, clearly outperforms deep learning methods even in large datasets.

A variety of multiagent systems methods has been proposed for forming cooperatives of interconnec... more A variety of multiagent systems methods has been proposed for forming cooperatives of interconnected agents representing electricity producers or consumers in the Smart Grid. One major problem that arises in this domain is assessing participating agents’ uncertainty, and correctly predicting their future behaviour regarding power consumption shifting actions. In this paper we adopt two stochastic filtering techniques, a Gaussian Process Filter and a Histogram Filter, and use these to effectively monitor the trustworthiness of agent statements regarding their final shifting actions. We incorporate these within a directly applicable scheme for providing electricity demand management services. Experiments were conducted on real-world consumption datasets from Kissamos, a municipality of Crete. Our results confirm that these techniques provide tangible benefits regarding enhanced consumption reduction performance, and increased financial gains for the cooperative.

A variety of multiagent systems methods has been proposed for forming cooperatives of interconnec... more A variety of multiagent systems methods has been proposed for forming cooperatives of interconnected agents representing electricity producers or consumers in the Smart Grid. One major problem that arises in this domain is assessing participating agents uncertainty, and correctly predicting their future behaviour. In this paper, we adopt two stochastic filtering techniques —the Unscented Kalman Filter equipped with Gaussian Processes, and the Histogram Filter— and use these to effectively monitor the trustworthiness of agent statements regarding their final actions. The methods are incorporated within a directly applicable scheme for providing electricity demand management services. Simulation results confirm that these techniques provide tangible benefits regarding enhanced consumption reduction performance, and increased financial gains.

Ο σκοπός αυτής της εργασίας είναι να εμβαθύνει την κατανόησή μας για την παρακμή του Μινωϊκού πολ... more Ο σκοπός αυτής της εργασίας είναι να εμβαθύνει την κατανόησή μας για την παρακμή του Μινωϊκού πολιτισμού, χρησιμοποιώντας ένα υπάρχον πολυπρακτορικό μοντέλο προσομοίωσης (ABM) για να μελετήσει σε ποιο βαθμό η κατακλυσμική έκρηξη του ηφαιστείου της Θήρας επηρέασε την κοινωνική εξέλιξη του Μινωϊκού πολιτισμού. Λαμβάνοντας υπόψη τη γεωργία ως κύρια παραγωγική δραστηριότητα για την διατήρηση του ανθρώπινου πληθυσμού, αξιολογούμε τις επιπτώσεις της έκρηξης του ηφαιστείου πάνω σε διαφορετικά μοντέλα κοινωνικής οργάνωσης, εστιάζοντας στην ευρύτερη περιοχή των Μαλίων της Κρήτης. Τα παραδείγματα κοινωνικής οργάνωσης που εξετάστηκαν είναι εμπνευσμένα από ένα μοντέλο αυτο-οργάνωσης μιας κοινότητας πρακτόρων καθώς και ιδέες από την εξελικτική θεωρία παιγνίων. Οι επιλογές παραμέτρων του πολυπρακτορικού μοντέλου βασίζονται σε αρχαιολογικές θεωρίες και ευρήματα, αλλά δεν είναι μεροληπτικές προς οποιαδήποτε συγκεκριμένη παραδοχή. Αποτελέσματα από διαφορετικά σενάρια προσομοίωσης επιδεικνύουν μια εν...

Tackling the decision-making problem faced by a prosumer (i.e., a producer that is simultaneously... more Tackling the decision-making problem faced by a prosumer (i.e., a producer that is simultaneously a consumer) when selling and buying energy in the emerging smart electricity grid, is of utmost importance for the economic profitability of such a business entity. In this paper, we model, for the first time, this problem as a factored Markov Decision Process. By so doing, we are able to represent the problem compactly, and provide an exact optimal solution via dynamic programming - notwithstanding its large size. Our model successfully captures the main aspects of the business decisions of a prosumer corresponding to a community microgrid of any size. Moreover, it includes appropriate sub-models for prosumer production and consumption prediction. Experimental simulations verify the effectiveness of our approach; and show that our exact value iteration solution matches that of a state-of-the-art method for stochastic planning in very large environments, while outperforming it in terms ...

We propose a novel technique for Active Malware Analysis (AMA) formalized as a Bayesian game betw... more We propose a novel technique for Active Malware Analysis (AMA) formalized as a Bayesian game between an analyzer agent and a malware agent, focusing on the decision making strategy for the analyzer. In our model, the analyzer performs an action on the system to trigger the malware into showing a malicious behavior, i.e., by activating its payload. The formalization is built upon the link between malware families and the notion of types in Bayesian games. A key point is the design of the utility function, which reflects the amount of uncertainty on the type of the adversary after the execution of an analyzer action. This allows us to devise an algorithm to play the game with the aim of minimizing the entropy of the analyzer’s belief at every stage of the game in a myopic fashion. Empirical evaluation indicates that our approach results in a significant improvement both in terms of learning speed and classification score when compared to other state-of-the-art AMA techniques. ACM Refe...

Simulation Modelling Practice and Theory, 2021

Proceedings of the 10th Hellenic Conference on Artificial Intelligence, 2018

In this article, we explore the computation of joint policies for autonomous agents to resolve co... more In this article, we explore the computation of joint policies for autonomous agents to resolve congestions problems in the air traffic management (ATM) domain. Agents, representing flights, have limited information about others' payoffs and preferences, and need to coordinate to achieve their tasks while adhering to operational constraints. We formalize the problem as a multiagent Markov decision process (MDP) towards deciding flight delays to resolve demand and capacity balance (DCB) problems in ATM. To this end, we present multiagent reinforcement learning methods that allow agents to interact and form own policies in coordination with others. Experimental study on real-world cases, confirms the effectiveness of our approach in resolving the demand-capacity balance problem.

Advances in Building Energy Research, 2020