Editors’ Introduction

Abstract

Learning theory is a research area involved in the study, design, and analysis of computer programs that are able to learn from past experiences. Over the last years the idea of learning system has independently emerged in a variety of fields: pattern recognition, theory of repeated games, machine learning, universal prediction and data compression, inductive inference, adaptive optimal control, computational psychology, and others. This is not surprising: on the one hand, the notion of adaptivity is an extremely attractive and powerful tool; on the other hand, the abstract phenomenon of learning is so complex that it is unlikely that a single theory will ever be able to explain it in a fully satisfactory way. Despite their apparent diversity, these learning models present deep connections whose study is an active research area in learning theory.