Computational Intelligence for Missing Data Imputation, Estimation, and Management: Knowledge Optimization Techniques

The issue of missing data imputation has been extensively explored in information engineering in recent years. Marwala's (systems engineering, U. of Witwatersrand, S. Africa) text examines various methods and technologies for estimating missing values given the observed data. An opening discussion of traditional missing data issues is followed by chapters covering a range of computational intelligence methods including the multi-layer perceptron model, radial basis functions, Gaussian mixture models, rough sets, support vector machines, decision trees, Fuzzy ARTMAP and extension neural networks. Their implementation in both the Bayesian and maximum-likelihood framework is explained. Marwala also discusses hybrid methods, particularly ensemble methods using more than one learning algorithm, and various optimization methods. Case studies are incorporated throughout to illustrate the concepts. For engineering, science, and statistical researchers and practitioners in aerospace, mechanical, civil, biomedical, and electrical engineering, and researchers in statistics and social science. Annotation c2009 Book News, Inc., Portland, OR (booknews.com)