Evolving Connectionist Systems: Methods and Applications in Bioinformatics, Brain Study and Intelligent Machines (Perspectives in Neural Computing)

Rating:
Summary: Real-time neural network with a host of applications
Review: I found this book to be a landmark contribution to the state-of-the-art in neural networks pardigm. It offers some exciting neural network topologies and a distinctly new kind of thinking -'local learning' in neural networks. The author Prof. Nik Kasabov deserves to be congratulated for writing this excellent book. His explanation throughout the book is very lucid and to the point. He introduces the concept of "evolving connectionism" in a succinct way. He included a rich assortment of connectionist methods, right from the scratch, with a clear exposition of the underlying training algorithms. The applications presented in the latter part of the book are as diverse as bioinformatics, financial engineering, speech recognition, brain study and image & video data processing. The authority with which these topics are presented speaks volumes of the enormous research work undertaken by Prof. Kasabov and his students. The references and extended glosary provided at the end are extremely useful to the reader. Another important aspect of this book is that it is suitable for all levels of readers such as student, researcher and practitioner. I started teaching some aspects of this book from this semester onwards. It is well received by the students. It must be in the shelves of those who look for the latest research in the area of neural networks. I enjoyed reading this book. Finally, if the phrase "real-time neural networks" is also added in the tag line (sub title) of the book, it could attract more users.

Rating:
Summary: An exceptional book for computational biologists
Review: This exceptional book provides a broad overview of the methods of extracting the knowledge (or in other words building model/system/theory) from the data in various areas: from information theory and artificial intelligence to genetics. It can be very useful for biologists, who wish to use modern computational methods for analysis of microarray data, regulatory networks, cancer, analysis of clinical trials, etc.

The first part (first seven chapters) of the book is devoted to the methods used in connectionists systems and here readers can find detailed description of the algorithms. In the second part (six chapters), which represents application of these methods, the book has a chapter devoted to the data analysis, modeling, and knowledge discovery in bioinformatics so it can be interesting for the biologists. This chapter describes how the neural network paradigm can be used in molecular biology and, in particular, for analysis in relatively new area -- microarray technology. The huge amount of data that were obtained in this area is still waiting for the efficient methods of knowledge extracting. In this chapter readers can also find the examples of using evolving connectionists learning systems for solving the problems of finding the patterns from DNA/RNA sequences, identification of intron/exon binding sites, gene profiling, protein structure prediction and dynamic cell modeling.

This excellent book is full of interesting examples, classification schemes, and figures.
Although this book will be more interesting for readers, which have been working in networking, it can be useful also for all researchers and students and any type of readers interesting in data analysis. This book is outstanding introduction for readers unfamiliar with the learning systems. The extended glossary and full-length reference list will help a lot for readers inexperienced in this area.