Artificial neural network is designed with the goal of building intelligent machines to solve complex perceptual problems, such as pattern classification, by mimicking the networks of real neurons in the human brain. To achieve better performance over a single neural network, ensembles construction with several neural networks has become a nutritious filed of research. This book presents the basics of ensemble construction and discusses several existing ensemble methods. A large number of benchmark problems from different application domains were considered for evaluating the methods. After the discussion of the existing methods three new different techniques of ensemble constructions are explained in three different chapters. Last two chapters of this book are to discuss the new techniques on a common ground and draw the conclusions of this work together with the outline of future directions of research opened by this work.