Stream Data Mining: Algorithms and Their Probabilistic Properties

Description

This book presents a unique approach to stream data mining. Unlike the vast majority of previous approaches, which are largely based on heuristics, it highlights methods and algorithms that are mathematically justified. First, it describes how to adapt static decision trees to accommodate data streams; in this regard, new splitting criteria are developed to guarantee that they are asymptotically equivalent to the classical batch tree. Moreover, new decision trees are designed, leading to the original concept of hybrid trees. In turn, nonparametric techniques based on Parzen kernels and orthogonal series are employed to address concept drift in the problem of non-stationary regressions and classification in a time-varying environment. Lastly, an extremely challenging problem that involves designing ensembles and automatically choosing their sizes is described and solved. Given its scope, the book is intended for a professional audience of researchers and practitioners who deal with stream data, e.g. in telecommunication, banking, and sensor networks. Front Matter ....Pages i-ix Introduction and Overview of the Main Results of the Book (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 1-10 Front Matter ....Pages 11-11 Basic Concepts of Data Stream Mining (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 13-33 Front Matter ....Pages 35-35 Decision Trees in Data Stream Mining (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 37-50 Splitting Criteria Based on the McDiarmid’s Theorem (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 51-62 Misclassification Error Impurity Measure (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 63-82 Splitting Criteria with the Bias Term (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 83-89 Hybrid Splitting Criteria (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 91-113 Front Matter ....Pages 115-115 Basic Concepts of Probabilistic Neural Networks (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 117-154 General Non-parametric Learning Procedure for Tracking Concept Drift (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 155-172 Nonparametric Regression Models for Data Streams Based on the Generalized Regression Neural Networks (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 173-244 Probabilistic Neural Networks for the Streaming Data Classification (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 245-277 Front Matter ....Pages 279-279 The General Procedure of Ensembles Construction in Data Stream Scenarios (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 281-286 Classification (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 287-308 Regression (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 309-322 Final Remarks and Challenging Problems (Leszek Rutkowski, Maciej Jaworski, Piotr Duda)....Pages 323-327 Back Matter ....Pages 329-330