Multi-objective optimization deals with the simultaneous optimization of two or more objectives which are normally in con?ict with each other. Since mul- objective optimization problems are relatively common in real-world appli- tions, this area has become a very popular research topic since the 1970s. However, the use of bio-inspired metaheuristics for solving multi-objective op- mization problems started in the mid-1980s and became popular until the mid- 1990s. Nevertheless, the e?ectiveness of multi-objective evolutionary algorithms has made them very popular in a variety of domains. Swarm intelligence refers to certain population-based metaheuristics that are inspired on the behavior of groups of entities (i.e., living beings) interacting locallywitheachotherandwiththeirenvironment.Suchinteractionsproducean emergentbehaviorthatismodelledinacomputerinordertosolveproblems.The two most popular metaheuristics within swarm intelligence are particle swarm optimization (which simulates a ?ock of birds seeking food) and ant colony optimization (which simulates the behavior of colonies of real ants that leave their nest looking for food). These two metaheuristics havebecome verypopular inthelastfewyears,andhavebeenwidelyusedinavarietyofoptimizationtasks, including some related to data mining and knowledge discovery in databases. However, such work has been mainly focused on single-objective optimization models. The use of multi-objective extensions of swarm intelligence techniques in data mining has been relatively scarce, in spite of their great potential, which constituted the main motivation to produce this book.
The purpose of this book is to collect contributions that are at the intersection of multi-objective optimization, swarm intelligence (specifically, particle swarm optimization and ant colony optimization) and data mining. Such a collection intends to illustrate the potential of multi-objective swarm intelligence techniques in data mining, with the aim of motivating more researchers in evolutionary computation and machine learning to do research in this field.
This volume consists of eleven chapters, including an introduction that provides the basic concepts of swarm intelligence techniques and a discussion of their use in data mining. Some of the research challenges that must be faced when using swarm intelligence techniques in data mining are also addressed. The rest of the chapters were contributed by leading researchers, and were organized according to the steps normally followed in Knowledge Discovery in Databases (KDD) (i.e., data preprocessing, data mining, and post processing).
We hope that this book becomes a valuable reference for those wishing to do research on the use of multi-objective swarm intelligence techniques in data mining and knowledge discovery in databases.
