Advances in intelligent data analysis
- 538 Seiten
- 19 Lesestunden
Inhaltsverzeichnis Learning covers a range of methodologies and techniques for intelligent data analysis, including statistical measures and linguistic model design. It discusses a "Top-Down and Prune" induction scheme for decision committees and explores mining clusters with association rules. The text delves into evolutionary computation for identifying strongly correlated variables in high-dimensional time-series data and examines biases in decision tree pruning strategies. Feature selection and retrospective pruning in hierarchical clustering are also addressed, alongside the discriminative power of input features in fuzzy models. Visualization techniques include monitoring human information processing through EEG analysis and knowledge-based visualization for spatial data mining. It introduces probabilistic topic maps for navigating large text collections and employs 3D visualizations for multidimensional data. Classification and clustering topics feature a decision tree algorithm for ordinal classification, Bayesian clustering for dynamic discovery, and nonparametric linear discriminant analysis. The text discusses supervised classification challenges and temporal pattern generation using hidden Markov models. Integration strategies include adjusted estimation for classifier combinations and reasoning about input-output modeling of dynamic systems. Applications range from intrusion detection and dairy industry pre
