Vicenc Torra Livres

Information fusion techniques and aggregation operators produce the most comprehensive, specific datum about an entity using data supplied from different sources, thus enabling us to reduce noise, increase accuracy, summarize and extract information, and make decisions. These techniques are applied in fields such as economics, biology and education, while in computer science they are particularly used in fields such as knowledge-based systems, robotics, and data mining. This book covers the underlying science and application issues related to aggregation operators, focusing on tools used in practical applications that involve numerical information. Starting with detailed introductions to information fusion and integration, measurement and probability theory, fuzzy sets, and functional equations, the authors then cover the following topics in detail: synthesis of judgements, fuzzy measures, weighted means and fuzzy integrals, indices and evaluation methods, model selection, and parameter extraction. The methods are illustrated with representative examples throughout, and there are extensive bibliographies and reading suggestions. The book is intended for graduate students, researchers, and practitioners such as engineers, computer scientists, statisticians and economists who use decision models and aggregation operators. The reader is assumed to have a nonspecialized background in mathematics.

The content includes invited talks and regular papers focusing on various aspects of fuzzy systems and machine learning. Topics cover RoboCup research in Japan, fuzzy integrals, and qualitative models for web quality evaluation. Other studies explore statistical and fuzzy neural network approaches for stock price forecasting, integration of wavelet time shift properties with support vector machines, and the weighting of training patterns in fuzzy rule-based classification systems. Innovative methods such as a quantum computing-based machine learning approach for radar emitter signal recognition and a multiagent system for modeling lower urinary tract control are discussed. The annotation also highlights mining diagnostic taxonomies from clinical databases and combining classifiers using Dempster’s rule for text categorization. Further, it examines assessment errors in LSP criteria, reliability of these criteria, and unsupervised aggregation via the Choquet integral for student evaluations. The preservation of fuzzy measure classes, decision modeling with the Choquet integral, and consensus measurements in multi-granular linguistic group decision-making are also addressed. The exploration of fuzzy granules in symbolic data analysis, learning procedures for estimating missing values in fuzzy preference relations, and decision-making in dynamic systems based on aggregated fuzzy preferences are included. Additionally, a comp

Information fusion is becoming a major requirement in data mining and knowledge discovery in databases. This book presents some recent fusion techniques that are currently in use in data mining, as well as data mining applications that use information fusion. Special focus of the book is on information fusion in preprocessing, model building and information extraction with various applications.