Data Mining in Finance presents a complete overview of major algorithmic approaches toward predictive data mining, counting statistical, neural networks, ruled-based, decision-tree, as well as fuzzy-logic methods, in addition to then examines the appropriateness of these approaches toward financial data mining. The book focuses specially on relational data mining, which is a learning technique able to learn more communicative rules than additional symbolic approaches.
RDM is thus improved suited for financial mining, for the reason that it is able toward make greater use of underlying domain knowledge. Relational data mining as well has a better capability on the way to give details the exposed rules - a capability critical for avoiding bogus patterns which unavoidably arise at what time the number of variables examined is extremely large.
Data Mining in Finance introduces a novel approach, combining relational data mining by means of the analysis of statistical implication of discovered rules. This reduces the search space as well as speeds up the algorithms. The book as well presents interactive as well as fuzzy-logic tools for `mining' the information as of the experts, promote reducing the search space.
