Rules are an accepted means of representing knowledge for virtually every domain. Traditional machine learning methods derive rules by exploring sets of examples using statistical or information theoretic techniques. Alternatively, rules can be discovered through methods of Evolutionary Computation such as genetic algorithms and learning classifier systems. In recent years, new models of learning classifier systems have been developed which have resulted in successful applications in a wide variety of domains (e.g., autonomous robotics, classification, knowledge discovery, modeling). These models have led to a resurgence of this area which for a certain period appeared almost at a dead end. This paper overviews the recent developments in learning classifier systems research, the new models, and the most interesting applications, suggesting some of the most relevant future research directions.
Learning classifier systems: new models, successful applications
LANZI, PIER LUCA;
2002-01-01
Abstract
Rules are an accepted means of representing knowledge for virtually every domain. Traditional machine learning methods derive rules by exploring sets of examples using statistical or information theoretic techniques. Alternatively, rules can be discovered through methods of Evolutionary Computation such as genetic algorithms and learning classifier systems. In recent years, new models of learning classifier systems have been developed which have resulted in successful applications in a wide variety of domains (e.g., autonomous robotics, classification, knowledge discovery, modeling). These models have led to a resurgence of this area which for a certain period appeared almost at a dead end. This paper overviews the recent developments in learning classifier systems research, the new models, and the most interesting applications, suggesting some of the most relevant future research directions.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
