ICA 2000 Proceedings. Proceedings of the Second International Workshop onIndependent Component Analysis and Blind Signal Separation. Held in Helsinki, Finland on June 19-22, 2000

摘要

Independent Component Analysis (ICA) has its roots in the active period ofartificial neural networks in the early and mid eighties, when many of the presently well-known topics and models were introduced. French researchers were the first to formulate the ICA problem in the context of blind source separation (BSS) and also coined the name. The concept was found to be useful in signal processing, and the emerging trend of using higher order statistical methods coincided with the interest in ICA. The basic instantaneous ICA model provided a statistical problem which, despite its apparent simplicity, was elegant and immediately useful in practice. The ICA2000 workshop program contains four invited papers by eminent scientists working on ICA, BSS, or closely related fields, as well as 99 technical papers. At the time of this writing, the number of registered participants is about 120. This promises to be a well focussed, high level workshop covering all the main aspects of ICA and blind source separation.