Different researchers hold different views of what the term meta-learning exactly means. The first part of this paper provides our own perspective view in which the goal is to build self-adaptive learners (i.e. learning algorithms that improve their bias dynamically through experience by accumulating meta-knowledge). The second part provides a survey of meta-learning as reported by the machine-learning literature. We find that, despite different views and research lines, a question remains constant: how can we exploit knowledge about learning (i.e. meta-knowledge) to improve the performance of learning algorithms? Clearly the answer to this question is key to the advancement of the field and continues being the subject of intensive research.

/pdf/a-perspective-view-and-survey-of-meta-learning-3z92vjfsfw.pdf

A perspective view and survey of meta-learning

Many researchers have demonstrated that the organizational design employed by an agent system can have a significant, quantitative effect on its performance characteristics A range of organizational strategies have emerged from this line of research, each with different strengths and weaknesses In this article we present a survey of the major organizational paradigms used in multi-agent systems These include hierarchies, holarchies, coalitions, teams, congregations, societies, federations, markets, and matrix organizations We will provide a description of each, discuss their advantages and disadvantages, and provide examples of how they may be instantiated and maintained This summary will facilitate the comparative evaluation of organizational styles, allowing designers to first recognize the spectrum of possibilities, and then guiding the selection of an appropriate organizational design for a particular domain and environment

https://cse.unl.edu/~lksoh/Classes/CSCE990AMAS_Spring13/Seminar04.pdf

A survey of multi-agent organizational paradigms

In Retsina, the authors have developed a distributed collection of software agents that cooperate asynchronously to perform goal-directed information retrieval and integration for supporting a variety of decision-making tasks. Examples for everyday organizational decision making and financial portfolio management demonstrate its effectiveness.

/pdf/distributed-intelligent-agents-414mvmm343.pdf

Distributed intelligent agents

The goal of the InfoSleuth project at MCC is to exploit and synthesize new technologies into a unified system that retrieves and processes information in an ever-changing network of information sources. InfoSleuth has its roots in the Carnot project at MCC, which specialized in integrating heterogeneous information bases. However, recent emerging technologies such as internetworking and the World Wide Web have significantly expanded the types, availability, and volume of data available to an information management system. Furthermore, in these new environments, there is no formal control over the registration of new information sources, and applications tend to be developed without complete knowledge of the resources that will be available when they are run. Federated database projects such as Carnot that do static data integration do not scale up and do not cope well with this ever-changing environment. On the other hand, recent Web technologies, based on keyword search engines, are scalable but, unlike federated databases, are incapable of accessing information based on concepts. In this experience paper, we describe the architecture, design, and implementation of a working version of InfoSleuth. We show how InfoSleuth integrates new technological developments such as agent technology, domain ontologies, brokerage, and internet computing, in support of mediated interoperation of data and services in a dynamic and open environment. We demonstrate the use of information brokering and domain ontologies as key elements for scalability.

InfoSleuth: agent-based semantic integration of information in open and dynamic environments

Scientific research and practice in multiagent systems focuses on constructing computational frameworks, principles, and models for how both small and large societies of intelligent, semiautonomous agents can interact effectively to achieve their goals. This article provides a personal view of the key application areas for cooperative multiagent systems, the major intellectual problems in building such systems, the underlying principles governing their design, and the major directions and challenges for future developments in this field.

/pdf/cooperative-multiagent-systems-a-personal-view-of-the-state-2g7pr0n623.pdf

Cooperative multiagent systems: a personal view of the state of the art

Two approaches to the problem of information-gathering, that may be characterised as distributed processing and distributed problem-solving, are contrasted. The former is characteristic of most existing information-gathering systems, and the latter is central to research in multi-agent systems. The features of complex information-carrying environments and the information-gathering task are examined, demonstrating both the utility of viewing information-gathering as distributed problem-solving and difficulties with viewing it as distributed processing. A new approach is proposed to information-gathering based on the distributed problem-solving paradigm and its attendant body of research in multi-agent systems and distributed artificial intelligence. This approach, called cooperative information-gathering, involves concurrent, asynchronous discovery and composition of information spread across a network of information servers. Top-level queries drive the creation of partially elaborated information-gathering plans, resulting in the employment of multiple semi-autonomous cooperative agents for the purpose of achieving goals and subgoals within those plans. The system as a whole satisfies, trading off solution quality and search cost while respecting user-imposed deadlines. Current work on distributed and agent-based approaches to information-gathering is also surveyed.

Cooperative information-gathering: a distributed problem-solving approach

Retrieval and Reasoning in Distributed Case Bases1

Achieving effective cooperation in a multi-agent system is a difficult problem for a number of reasons such as limited and possibly out-dated views of activities of other agents and uncertainty about the outcomes of interacting non-local tasks. In this paper, we present a learning system called COLLAGE, that endows the agents with the capability to learn how to choose the most appropriate coordination strategy from a set of available coordination strategies. COLLAGE relies on meta-level information about agents' problem solving situations to guide them towards a suitable choice for a coordination strategy. We present empirical results that strongly indicate the effectiveness of the learning algorithm.

/pdf/learning-situation-specific-coordination-in-cooperative-399a6o9swp.pdf

Learning Situation-Specific Coordination in Cooperative Multi-agent Systems

When we design multl-agent systems for realistic, worth-oriented environments, solving the coordination problems they present involves understanding the intricate and sophisticated interplay between the domain and the various system components. In this paper, we examine this interplay using ",’:E,~I.~ and the GPGP (Generalized Partial Global Planning) approach as a tool for quickly exploring multlple alternative organizations/coordination strategies. We illustrate the flexibility of "r.~M~-style task structures for representing interesting multi-criteria coordination problems, and the use of a new grammar-based generation tool to allow quicker task structure experimentation. Using Distributed Data Processing as an example domain, we show how different coordination algorithms of varying sophistication give rise to subtle effects in functionallystructured agent organizations with interrelationships between functions. The experiments presented in our paper serve to illustrate our claim about the generality and flexibility of the I’~:.~!~ representation for coordination problems and the grammar-based generators for modeling domain semantics and experimenting with different aspects of such domains.

/pdf/exploring-organizational-designs-with-t-1ems-a-case-study-of-1tcn7fg59c.pdf

Exploring Organizational Designs with T.1EMS: A Case Study of Distributed Data Processing*

Achieving effective cooperation in a multi-agent system is a difficult problem for a number of reasons such as limited and possibly out-dated views of activities of other agents and uncertainty about the outcomes of interacting non-local tasks. In this paper, we present a learning algorithm that endows agents with the capability to choose the appropriate coordination algorithm from a set of available coordination algorithms based on meta-level information about their problem solving situations. We present empirical results that strongly indicate the effectiveness of the learning algorithm.

/pdf/the-use-of-meta-level-information-in-learning-situation-34yq97lz4z.pdf

M. V. Nagendra Prasad

Papers

Cooperative information-gathering: a distributed problem-solving approach

Retrieval and Reasoning in Distributed Case Bases1

Learning Situation-Specific Coordination in Cooperative Multi-agent Systems

Exploring Organizational Designs with T.1EMS: A Case Study of Distributed Data Processing*

The Use of Meta-level Information in Learning Situation-Specific Coordination