Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

An Introduction to MultiAgent Systems.

One of the most important design problems for multi-UAV (Unmanned Air Vehicle) systems is the communication which is crucial for cooperation and collaboration between the UAVs. If all UAVs are directly connected to an infrastructure, such as a ground base or a satellite, the communication between UAVs can be realized through the in-frastructure. However, this infrastructure based communication architecture restricts the capabilities of the multi-UAV systems. Ad-hoc networking between UAVs can solve the problems arising from a fully infrastructure based UAV networks. In this paper, Flying Ad-Hoc Networks (FANETs) are surveyed which is an ad hoc network connecting the UAVs. The differences between FANETs, MANETs (Mobile Ad-hoc Networks) and VANETs (Vehicle Ad-Hoc Networks) are clarified first, and then the main FANET design challenges are introduced. Along with the existing FANET protocols, open research issues are also discussed.

Flying Ad-Hoc Networks (FANETs)

Network anomaly detection is an important and dynamic research area. Many network intrusion detection methods and systems (NIDS) have been proposed in the literature. In this paper, we provide a structured and comprehensive overview of various facets of network anomaly detection so that a researcher can become quickly familiar with every aspect of network anomaly detection. We present attacks normally encountered by network intrusion detection systems. We categorize existing network anomaly detection methods and systems based on the underlying computational techniques used. Within this framework, we briefly describe and compare a large number of network anomaly detection methods and systems. In addition, we also discuss tools that can be used by network defenders and datasets that researchers in network anomaly detection can use. We also highlight research directions in network anomaly detection.

Network Anomaly Detection: Methods, Systems and Tools

This survey provides a structured and comprehensive overview of research on security and privacy in computer and communication networks that use game-theoretic approaches. We present a selected set of works to highlight the application of game theory in addressing different forms of security and privacy problems in computer networks and mobile applications. We organize the presented works in six main categories: security of the physical and MAC layers, security of self-organizing networks, intrusion detection systems, anonymity and privacy, economics of network security, and cryptography. In each category, we identify security problems, players, and game models. We summarize the main results of selected works, such as equilibrium analysis and security mechanism designs. In addition, we provide a discussion on the advantages, drawbacks, and future direction of using game theory in this field. In this survey, our goal is to instill in the reader an enhanced understanding of different research approaches in applying game-theoretic methods to network security. This survey can also help researchers from various fields develop game-theoretic solutions to current and emerging security problems in computer networking.

/pdf/game-theory-meets-network-security-and-privacy-1r8g3nc948.pdf

Game theory meets network security and privacy

We address the problem of multi-UAV-based surveillance in complex urban environments with occlusions. The problem consists of controlling the flight of UAVs with on-board cameras so that the coverage and recency of the information about the designated area is maximized. In contrast to the existing work, sensing constraints due to occlusions and UAV motion constraints are modeled realistically and taken into account. We propose a novel \emph{occlusion-aware} surveillance algorithm based on a decomposition of the surveillance problem into a variant of the 3D art gallery problem and an instance of traveling salesman problem for Dubins vehicles. The algorithm is evaluated on the high-fidelity \textsc{AgentFly} UAV simulation testbed which accurately models all constraints and effects involved. The results confirm the importance of occlusion-aware flight path planning, in particular in the case of narrow-street areas and low UAV flight altitudes.

Autonomous UAV Surveillance in Complex Urban Environments

Invited Talks.- Making Agents Acceptable to People.- Coalition Formation: Towards Feasible Solutions Abstract of a Key-Note Speech.- Coalition Task Support Using I-X and ?I-N-C-A?.- Formal Methods.- Towards Motivation-Based Decisions for Worth Goals.- Modal Structure for Agents Interaction Based on Concurrent Actions.- A Multi-agent Modal Language for Concurrency with Non-communicating Agents.- Self-Synchronization of Cooperative Agents in a Distributed Environment.- MIP-Nets: A Compositional Model of Multiagent Interaction.- Social Knowledge & Meta-Reasoning.- Calibrating Collective Commitments.- Abstract Architecture for Meta-reasoning in Multi-agent Systems.- Balancing Individual Capabilities and Social Peer Pressure for Role Adoption.- From Social Agents to Multi-agent Systems: Preliminary Report.- Negotiation & Policies.- DAML-Based Policy Enforcement for Semantic Data Transformation and Filtering in Multi-agent Systems.- Architectures for Negotiating Agents.- RIO : Roles, Interactions and Organizations.- Conversation Mining in Multi-agent Systems.- Ontologies & Languages.- The Knowledge Market: Agent-Mediated Knowledge Sharing.- Ontology of Cooperating Agents by Means of Knowledge Components.- Mapping between Ontologies in Agent Communication.- A Social ACL Semantics by Deontic Constraints.- A Formal Specification Language for Agent Conversations.- Planning.- Framework for Multi-agent Planning Based on Hybrid Automata.- Multi-agent System for Resource Allocation and Scheduling.- Towards Autonomous Decision Making in Multi-agent Environments Using Fuzzy Logic.- Towards an Object Oriented Implementation of Belief-Goal-Role Multi-agent Systems.- Coalitions.- Fuzzy Coalition Formation among Rational Cooperative Agents.- Multi-agent Simulation of Work Teams.- Multi-agent Knowledge Logistics System "KSNet": Implementation and Case Study for Coalition Operations.- Evolution & Emergent Behavior.- Learning User Preferences for Multi-attribute Negotiation: An Evolutionary Approach.- A Model of Co-evolution in Multi-agent System.- Emergence of Specialized Behavior in a Pursuit-Evasion Game.- On a Dynamical Analysis of Reinforcement Learning in Games: Emergence of Occam's Razor.- Forgiveness in Strategies in Noisy Multi-agent Environments.- Platforms.- An Unified Framework for Programming Autonomous, Intelligent and Mobile Agents.- Tailoring an Agent Architecture to a Flexible Platform Suitable for Cooperative Robotics.- Airports for Agents: An Open MAS Infrastructure for Mobile Agents.- Beyond Prototyping in the Factory of Agents.- Agent Oriented Software Engineering with INGENIAS.- Protocols.- Requirement Analysis for Interaction Protocols.- Engineering a Protocol Server Using Strategy-Agents.- Refinement of Open Protocols for Modelling and Analysis of Complex Interactions in Multi-agent Systems.- Security.- Biological Approach to System Information Security (BASIS): A Multi-agent Approach to Information Security.- Adaptive Agents Applied to Intrusion Detection.- Communication Security in Multi-agent Systems.- Teamwork of Hackers-Agents: Modeling and Simulation of Coordinated Distributed Attacks on Computer Networks.- Real-Time & Synchronization.- Formal Modeling of Dynamic Environments for Real-Time Agents.- Deliberative Server for Real-Time Agents.- Regional Synchronization for Simultaneous Actions in Situated Multi-agent Systems.- A Multi-agent System for Dynamic Network Reconfiguration.- Industrial Applications.- A Highly Distributed Intelligent Multi-agent Architecture for Industrial Automation.- The Cambridge Packing Cell - A Holonic Enterprise Demonstrator.- Towards Autonomy, Self-Organisation and Learning in Holonic Manufacturing.- An Agent-Based Personalized Producer/Consumer Scenario.- Application of Intelligent Agents in Power Industry: Promises and Complex Issues.- E-business & Virtual Enterprises.- Brokering in Electronic Insurance Markets.- Modelling Electronic Organizations.- The Use of Adaptive Negotiation by a Shopping Agent in Agent-Mediated Electronic Commerce.- Agent Interaction Protocols for the Selection of Partners for Virtual Enterprises.- Web & Mobile Agents.- A Multiagent-Based Peer-to-Peer Network in Java for Distributed Spam Filtering.- Engineering Web Service Invocations from Agent Systems.- A Component Based Multi-agent Architecture to Support Mobile Business Processes.- Code Complexity Metrics for Mobile Agents Implemented with Aspect/J(TM).

Multi-Agent Systems and Applications III

Autonomous UAV Surveillance in Complex Urban Environments.

Autonomous Agents and Multi-Agent Systems are, in the judgment of many observers, the most significant new paradigm for software modelling and development to emerge from Computer Science in the last two decades. Autonomous Agents are computer programs that are able to decide between different methods of achieving their programmed goals. This is in contrast to the scripted, predefined behaviour that a traditional program, for example a Unix Daemon, would exhibit, and means that Autonomous Agents can cope with dynamic environments, can be developed as a quick solution to complex problems, and can tailor their behaviour to individual users in a way that traditional software cannot.

Multi-Agent Systems are computational environments in which different programs, possibly developed in isolation, can co-ordinate their behaviours to achieve their goals. The technology of Multi-Agent Systems is, therefore, particularly applicable to modern operational environments like e-business, ubiquitous computing, and, of course, the Internet.

These two topics of research are tightly coupled. For a Multi-Agent System to be fully implemented, it is often necessary to utilize the technology of Autonomous Agents. For an Agent to act with true autonomy in a realistic, modern setting, it must often be able to reason about other Agents and co-ordinate its behaviour with them. It is this insight that has led to the development of the AAMAS community and the spectacularly successful series of conferences that have been running for the last for years.

The AAMAS community has long recognised [Wooldridge & Jennings 1995] that the significance of the abstractions, techniques, tools, and technologies developed by the research community worldwide will only be recognised if they have practical applications in the real world of science, technology, education, healthcare, business, and commerce. The paradigm of AAMAS and available agent technologies has reached a significant level of maturity, and they are now widely regarded as ready for wider adoption. The purpose of the First Industry Track of the Autonomous Agent and Multi-Agent Systems conference is to provide a forum that will bring real world applications that are being developed by teams internationally to the attention of the AAMAS community and the wider world.

Why is this important? We have identified several motivations:

•By disseminating the news of successful application of AAMAS technology we hope to provide researchers with clear evidence of the success and usefulness of particular techniques.

•Feedback from attempts to use AAMAS technology is valuable in the formation of the AAMAS research agenda.

•A forum for discussion of industrial and application-orientated concerns has been lacking in the community to date; by providing one, we hope that practitioners will be able to share their experiences and, therefore, accelerate the uptake of AAMAS technology.

•Finally, we expect that, by providing an industry track, we will encourage more participation in AAMAS by practitioners, and that this will lead to a wider understanding of the importance of AAMAS technology in commercial organisations.

Fifteen papers that discuss the application of AAMAS technology are presented in this collection. They were selected by a refereeing process that aimed to find work that made use of the particular properties of AAMAS systems and had actually been used "in anger." The ratio of submitted to accepted papers was higher than 2:1; that is, more papers were excluded from presentation than were accepted.

The papers that have been selected by the referees and programme committee for the track allow the identification of some trends in the development and application of AAMAS technology. Four out of fifteen papers are in the field of transport, traffic, and logistics. Natural problem decomposition, geographical distribution, and requirements for autonomous decision making make this domain particularly suitable for AAMAS technology.

A further four papers are focused on Aerospace applications. In this domain, the technological adventurousness of a highly competitive and demanding industry may explain the work presented. However, the use of AAMAS technology to provide autonomy for systems that cannot be easily or cheaply supervised by human intervention seems to be another clear driver.

Three papers focus on using AAMAS technology for manufacturing applications. Here, the drivers seem to have been to use Agent Autonomy or co-ordination techniques to provide cost savings, realised by removing large numbers of repetitive management and control tasks.

The track also features papers which describe the use of AAMAS technology for electricity network management and to provide a training system for naval personnel. Both of these papers point to cost savings realised by automating decision making processes.

Just as we can see trends in the papers that have been selected for publication, it's also interesting to point to missing areas. There are no papers from promising domains for AAMAS adoption including telecommunications, internet, pharmaceuticals, eGovernment, or healthcare. Similarl,y the submission rate from small companies and start-ups was expected to be somewhat higher. These are areas that have been explored extensively in the past [Luck et-al 2003] so it is important for the AAMAS community to investigate why the research that has been performed has not matured into published case studies at this time.

Although the collection is diverse, all the papers share two common traits. Every paper is the result of the efforts of a team that has been prepared to take the risk of adopting a new technology in a high pressure environment, and every paper summarises what happens when talented people take a risk and that risk pays off. We salute all these pioneers, and look forward to the presentation of the technical program and the discussions that are bound to result.

Proceedings of the fourth international joint conference on Autonomous agents and multiagent systems

Individual anomaly-detection methods for monitoring computer network traffic have relatively high error rates. An agent-based trust-modeling system fuses anomaly data and progressively improves classification to achieve acceptable error rates.

Michal Pechoucek

Papers

Autonomous UAV Surveillance in Complex Urban Environments

Multi-Agent Systems and Applications III

Autonomous UAV Surveillance in Complex Urban Environments.

Proceedings of the fourth international joint conference on Autonomous agents and multiagent systems

Adaptive Multiagent System for Network Traffic Monitoring