Loading...

Data Stream Classification by Considering Concept Drift Using Evolutionary Algorithms

Karimi, Zohre | 2010

522 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 40167 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Abolhassani, Hassan
  7. Abstract:
  8. Today many applications produce data stream that grow continuously over time. Classification is an important task in mining this data. One of the major challenges in classification of data streams is that the underlying concept of data may change over time; this needs updating the classification model. Many of traditional classification algorithms are adapted for classifying data streams. Evolutionary algorithms are stochastic methods that have been successfully applied to a wide range of optimization problems. Harmony search is a new evolutionary algorithm that convergence to global optimum and can be applied to discrete and continuous data. The evolutionary nature of classifying data streams that matched on problem solving method in evolutionary algorithms, restricted research on applying evolutionary algorithms to data stream classification and benefits of harmony search are our motivations on using harmony search for data stream classification. In this thesis, a new classifier based on harmony search is proposed. An incremental model of this classifier is also proposed for classifying data streams that has the following advantages: appropriate data selection for training, flexibility in response time and stability in noisy data. In our experiments, the accuracy of proposed methods is compared with non-evolutionary classifiers on some standard datasets. The synthetic and real world data are used to validate the accuracy, memory usage and training time of Incremental model with significantly improved results compared to non-incremental harmony based classifier in the classification of data streams. The time and memory complexity and convergence proof of proposed methods to global optimum is also presented.

  9. Keywords:
  10. Harmony Search Algorithm ; Data Stream ; Evolutionary Algorithm ; Classification ; Concept Drift

 Digital Object List

 Bookmark

No TOC