Data Mining Methods

DATA MINING METHODS, Second Edition discusses both theoretical foundation and practical applications of datamining in a web field including banking, e-commerce, medicine, engineering and management. This book starts byintroducing data and information, basic data type, data category and applications [...]
Visa längre beskrivning

DATA MINING METHODS, Second Edition discusses both theoretical foundation and practical applications of datamining in a web field including banking, e-commerce, medicine, engineering and management. This book starts byintroducing data and information, basic data type, data category and applications of data mining. The second chapterbriefly reviews data visualization technology and importance in data mining. Fundamentals of probability and statisticsare discussed in chapter 3, and novel algorithm for sample covariants are derived. The next two chapters give an indepthand useful discussion of data warehousing and OLAP. Decision trees are clearly explained and a new tabularmethod for decision tree building is discussed. The chapter on association rules discusses popular algorithms andcompares various algorithms in summary table form. An interesting application of genetic algorithm is introduced inthe next chapter. Foundations of neural networks are built from scratch and the back propagation algorithm is derivedin the appendix. Popular clustering algorithm is discussed in the next chapter. The web mining chapter generalizes thepage rank metric in multiple ways.A geometric derivation of SDM appear next and summary table in table form isgiven. LSI indexing for IRN extension is discussed next. The book ends with a thorough discussion of text miningmetrics and gives latest research directions in text mining.KEY FEATURES:iC* "Application sections" that demonstrate the usefulness of models presented in each chapteriC* Large number of URL links to software on the net using which readers can build various data mining modelson their owniC* Extensive reference section at the end of each chapter, with 300+ research publications citediC* More than 250 exercises (true/false, multiple choice, computer exercises)