Data Compression for Data Mining Algorithms tackles the important problems in the design of more efficient data mining algorithms by way of data compression techniques and provides the first systematic and comprehensive description of the relationships between data compression mechanisms and the computations involved in data mining algorithms. Data mining algorithms are powerful analytical techniques used across various disciplines, including business, engineering, and science. However, in the big data era, tasks such as association rule mining and classification often require multiple scans of databases, while clustering and outlier detection methods typically depend on Euclidean distance for similarity measures, leading to high computational costs.
Data Compression for Data Mining Algorithms addresses these challenges by focusing on the scalarization of data mining algorithms, leveraging data compression techniques to reduce dataset sizes and applying information theory principles to minimize computations involved in tasks such as feature selection and similarity computation. The book features the latest developments in both lossless and lossy data compression methods and provides a comprehensive exposition of data compression methods for data mining algorithm design from multiple points of view.
Key discussions include Huffman coding, scalar and vector quantization, transforms, subbands, wavelet-based compression for scalable algorithms, and the role of neural networks, particularly deep learning, in feature selection and dimensionality reduction. The book’s contents are well-balanced for both theoretical analysis and real-world applications, and the chapters are well organized to compose a solid overview of the data compression techniques for data mining. To provide the reader with a more complete understanding of the material, projects and problems solved with Python are interspersed throughout the text.
1. Introduction - p. xix 2. Part I. Foundation - p. P1 3. Chapter 1. Overview and contributions - p. 3 4. Chapter 2. Introduction to data compression methods - p. 13 5. Chapter 3. Introduction to data mining algorithms - p. 29 6. Part II. Association rule mining - p. P2 7. Chapter 4. Arithmetic coding for maximal frequent itemset mining - p. 51 8. Chapter 5. Data compression for association rule mining - p. 65 9. Part III. Classification - p. P3 10. Chapter 6. Feature selection for decision tree construction - p. 83 11. Chapter 7. Principal component analysis for decision tree classification - p. 117 12. Chapter 8. Neural networks for non-linear dimensionality reduction - p. 135 13. Chapter 9. Dictionary-based coding techniques for support vector machines - p. 151 14. Chapter 10. Quantization for support vector machines - p. 171 15. Chapter 11. Deep learning for regression and feature selection - p. 189 16. Chapter 12. Wavelet based denoising for autoregression - p. 205 17. Part IV. Clustering and outlier detection - p. P4 18. Chapter 13. A sparse data representation for clustering - p. 227 19. Chapter 14. Dictionary-based coding compression for clustering - p. 251 20. Chapter 15. Compression based nearest neighbor clustering for outlier detection - p. 273 21. Chapter 16. Huffman coding for outlier detection - p. 291 22. Chapter 17. Arithmetic coding for outlier detection - p. 311 23. Index - p. 337
Data Compression for Data Mining Algorithms tackles the important problems in the design of more efficient data mining algorithms by way of data compression techniques and provides the first systematic and comprehensive description of the relationships between data compression mechanisms and the computations involved in data mining algorithms. Data mining algorithms are powerful analytical techniques used across various disciplines, including business, engineering, and science. However, in the big data era, tasks such as association rule mining and classification often require multiple scans of databases, while clustering and outlier detection methods typically depend on Euclidean distance for similarity measures, leading to high computational costs.
Data Compression for Data Mining Algorithms addresses these challenges by focusing on the scalarization of data mining algorithms, leveraging data compression techniques to reduce dataset sizes and applying information theory principles to minimize computations involved in tasks such as feature selection and similarity computation. The book features the latest developments in both lossless and lossy data compression methods and provides a comprehensive exposition of data compression methods for data mining algorithm design from multiple points of view.
Key discussions include Huffman coding, scalar and vector quantization, transforms, subbands, wavelet-based compression for scalable algorithms, and the role of neural networks, particularly deep learning, in feature selection and dimensionality reduction. The book’s contents are well-balanced for both theoretical analysis and real-world applications, and the chapters are well organized to compose a solid overview of the data compression techniques for data mining. To provide the reader with a more complete understanding of the material, projects and problems solved with Python are interspersed throughout the text.
1. Introduction - p. xix 2. Part I. Foundation - p. P1 3. Chapter 1. Overview and contributions - p. 3 4. Chapter 2. Introduction to data compression methods - p. 13 5. Chapter 3. Introduction to data mining algorithms - p. 29 6. Part II. Association rule mining - p. P2 7. Chapter 4. Arithmetic coding for maximal frequent itemset mining - p. 51 8. Chapter 5. Data compression for association rule mining - p. 65 9. Part III. Classification - p. P3 10. Chapter 6. Feature selection for decision tree construction - p. 83 11. Chapter 7. Principal component analysis for decision tree classification - p. 117 12. Chapter 8. Neural networks for non-linear dimensionality reduction - p. 135 13. Chapter 9. Dictionary-based coding techniques for support vector machines - p. 151 14. Chapter 10. Quantization for support vector machines - p. 171 15. Chapter 11. Deep learning for regression and feature selection - p. 189 16. Chapter 12. Wavelet based denoising for autoregression - p. 205 17. Part IV. Clustering and outlier detection - p. P4 18. Chapter 13. A sparse data representation for clustering - p. 227 19. Chapter 14. Dictionary-based coding compression for clustering - p. 251 20. Chapter 15. Compression based nearest neighbor clustering for outlier detection - p. 273 21. Chapter 16. Huffman coding for outlier detection - p. 291 22. Chapter 17. Arithmetic coding for outlier detection - p. 311 23. Index - p. 337
