The Intelligent Data Engineering and Automated Learning (IDEAL) conf- ence series began in 1998 in Hong Kong, when the world started to experience information and data explosion and to demand for better, intelligent meth- ologies and techniques. It has since developed, enjoyed success in recent years, and become a unique annual international forum dedicated to emerging topics and technologies in intelligent data analysis and mining, knowledge discovery, automated learning and agent technology, as well as interdisciplinary appli- tions, especially bioinformatics. These techniques are common and applicable to many ?elds. The multidisciplinary nature of research nowadays is pushing the boundaries and one of the principal aims of the IDEAL conference is to p- mote interactions and collaborations between disciplines, which are bene?cial and bringing fruitful solutions. This volume of Lecture Notes in Computer Science contains accepted papers presented at IDEAL 2004, held in Exeter, UK, August 25-27, 2004. The conf- ence received 272 submissions from all over the world, which were subsequently refereed by the ProgramCommittee. Among them 124 high-quality papers were accepted and included in the proceedings. IDEAL 2004 enjoyed outstanding keynote talks by distinguished guest speakers, Jim Austin, Mark Girolami, Ross King, Lei Xu and Robert Esnouf. This year IDEAL also teamed up with three international journals, namely the International Journal of Neural Systems, the Journal of Mathematical M- elling and Algorithms, and Neural Computing & Applications. Three special issues on Bioinformatics, Learning Algorithms, and Neural Networks & Data Mining, respectively, have been scheduled for selected papers from IDEAL 2004.

Independent Components Analysis (ICA) is an important tool for modeling and understanding empirical data sets. Belonging to the class of general linear models, it is a method of separating out independent sources from linearly mixed data. ICA provides a better decomposition than other well-known models such as principal component analysis. This self-contained book contains a structured series of edited papers by leading researchers in the field and includes an extensive introduction to ICA. It reviews the major theoretical bases from a modern perspective, surveys current developments, and describes many case studies of applications in detail. Applications include biomedical examples, signal and image denoising, and mobile communications. The book discusses ICA within the framework of general linear models, but it also compares it to other paradigms such as neural network and graphical modeling methods.