Small Sample Modelling Based on Deep and Broad Forest Regression: Theory and Industrial Application delves into tree-structured methods in the industrial sector, encompassing classical ensemble learning, tree-structured deep forest classification, and broad learning systems with neural networks. It introduces an innovative deep/broad learning algorithm for small-sample industrial modeling tasks. The book is divided into two parts: methodology and practical application in dioxin emission modeling. Methodology sections include Preliminaries, Deep Forest Regression, Broad Forest Regression, and Fuzzy Forest Regression. The application part focuses on modeling dioxin emissions in municipal solid waste incineration. Throughout, various tree-structured strategies are presented, and the authors provide software systems for validating these methods. This book is suitable for advanced undergraduates, graduate engineering students, and practicing engineers looking for self-study resources.
1. Title of Book - p. i 2. Cover image - p. i 3. Title page - p. i 4. Table of Contents - p. iii 5. Emerging Methodologies and Applications in Modelling, Identification, and Control series: - p. ii 6. Copyright - p. iv 7. About the series editor - p. vii 8. Preface - p. ix 9. 1. Data acquisition and characteristics in industrial processes∗ - p. 1 10. 2. Modeling and adaptation for dynamic industrial data∗ - p. 11 11. 2.1 Relevant method description for modeling small sample data - p. 11 12. 2.2 Concept drift detection for regression models - p. 35 13. 2.3 Online updates for regression models - p. 47 14. 3. Deep forest regression for industrial modeling∗ - p. 65 15. 3.1 Basic deep forest regression method - p. 65 16. 3.2 DFR based on cross-layer fully connection - p. 75 17. 3.3 Simulation results - p. 83 18. 3.4 Conclusions - p. 99 19. 4. Broad forest regression for industrial modeling∗ - p. 101 20. 5. Fuzzy forest regression for industrial modeling∗ - p. 131 21. 5.1 Fuzzy forest regression strategy - p. 131 22. 5.2 Algorithm implementation - p. 133 23. 5.3 Simulation results - p. 146 24. 5.4 Conclusion - p. 154 25. 6. Deep forest regression based on feature reduction and feature enhancement∗ - p. 157 26. 6.1 Dioxin emission soft-sensing description for municipal solid waste incineration - p. 157 27. 6.2 Data-filling processing - p. 165 28. 6.3 DXN soft sensing - p. 184 29. 6.4 Conclusions - p. 197 30. 7. Simplified deep forest regression with combined feature selection and residual error fitting∗ - p. 203 31. 8. Online fuzzy broad forest regression∗ - p. 263 32. 8.1 Ensemble fuzzy forest regression for mechanism modeling particle matter (PM) - p. 263 33. 8.2 Online fuzzy broad forest regression for soft-sensing dioxin (DXN) - p. 285 34. 8.3 Robust online fuzzy broad forest regression for soft-sensing DXN - p. 307 35. 8.4 Application verification results - p. 324 36. 8.5 Conclusion - p. 330 37. Index - p. 335
Small Sample Modelling Based on Deep and Broad Forest Regression: Theory and Industrial Application delves into tree-structured methods in the industrial sector, encompassing classical ensemble learning, tree-structured deep forest classification, and broad learning systems with neural networks. It introduces an innovative deep/broad learning algorithm for small-sample industrial modeling tasks. The book is divided into two parts: methodology and practical application in dioxin emission modeling. Methodology sections include Preliminaries, Deep Forest Regression, Broad Forest Regression, and Fuzzy Forest Regression. The application part focuses on modeling dioxin emissions in municipal solid waste incineration. Throughout, various tree-structured strategies are presented, and the authors provide software systems for validating these methods. This book is suitable for advanced undergraduates, graduate engineering students, and practicing engineers looking for self-study resources.
1. Title of Book - p. i 2. Cover image - p. i 3. Title page - p. i 4. Table of Contents - p. iii 5. Emerging Methodologies and Applications in Modelling, Identification, and Control series: - p. ii 6. Copyright - p. iv 7. About the series editor - p. vii 8. Preface - p. ix 9. 1. Data acquisition and characteristics in industrial processes∗ - p. 1 10. 2. Modeling and adaptation for dynamic industrial data∗ - p. 11 11. 2.1 Relevant method description for modeling small sample data - p. 11 12. 2.2 Concept drift detection for regression models - p. 35 13. 2.3 Online updates for regression models - p. 47 14. 3. Deep forest regression for industrial modeling∗ - p. 65 15. 3.1 Basic deep forest regression method - p. 65 16. 3.2 DFR based on cross-layer fully connection - p. 75 17. 3.3 Simulation results - p. 83 18. 3.4 Conclusions - p. 99 19. 4. Broad forest regression for industrial modeling∗ - p. 101 20. 5. Fuzzy forest regression for industrial modeling∗ - p. 131 21. 5.1 Fuzzy forest regression strategy - p. 131 22. 5.2 Algorithm implementation - p. 133 23. 5.3 Simulation results - p. 146 24. 5.4 Conclusion - p. 154 25. 6. Deep forest regression based on feature reduction and feature enhancement∗ - p. 157 26. 6.1 Dioxin emission soft-sensing description for municipal solid waste incineration - p. 157 27. 6.2 Data-filling processing - p. 165 28. 6.3 DXN soft sensing - p. 184 29. 6.4 Conclusions - p. 197 30. 7. Simplified deep forest regression with combined feature selection and residual error fitting∗ - p. 203 31. 8. Online fuzzy broad forest regression∗ - p. 263 32. 8.1 Ensemble fuzzy forest regression for mechanism modeling particle matter (PM) - p. 263 33. 8.2 Online fuzzy broad forest regression for soft-sensing dioxin (DXN) - p. 285 34. 8.3 Robust online fuzzy broad forest regression for soft-sensing DXN - p. 307 35. 8.4 Application verification results - p. 324 36. 8.5 Conclusion - p. 330 37. Index - p. 335
