Table of Contents

Weather Prediction Using Case-Based Reasoning and Fuzzy Set Theory

Hansen, B. K., 2000: Weather Prediction Using Case-Based Reasoning and Fuzzy Set Theory, Master of Computer Science Thesis, Technical University of Nova Scotia, Halifax, Nova Scotia, Canada.

Table of Contents

List of Figures

List of Abbreviations and Symbols

Acknowledgments

1. Introduction

1.1 Hypothesis
1.2 Thesis structure
1.3 Case-based reasoning

1.3.1 Approaches to case-based reasoning 1.3.2 Challenges for case-based reasoning 1.3.3 Retrieval of similar cases

1.4 Fuzzy logic

1.4.1 Fuzzy logic enables retrieval of similar cases

1.5 Weather prediction

1.5.1 Airport weather prediction problem 1.5.2 State of the art of AI in weather prediction 1.5.3 Analog forecasting: An empirical weather prediction technique that depends on retrieval of similar cases

2. Literature Survey

2.1 Resources for case-based reasoning

2.1.1 Large databases of cases 2.1.2 Domain knowledge about similarity 2.1.3 Commercial software tools

2.2 Fuzzy logic and case-based reasoning

2.2.1 Fuzzy CBR formalism 2.2.2 Numerous conditions and partial matching 2.2.3 Flexible similarity-measuring framework 2.2.4 Numbers and words

2.3 Foundation for fuzzy k-nearest neighbors technique

2.3.1 k-nearest neighbors technique 2.3.2 Fuzzy k-nearest neighbors technique 2.3.3 Weather situations are not prototypical 2.3.4 Properties of the fuzzy k-nn technique

2.4 Applications that use fuzzy k-nn techniques

2.4.1 Weather prediction 2.4.2 Mergers and acquisitions 2.4.3 Residential property valuation 2.4.4 Cash flow forecasting 2.4.5 Shoe fashion database retrieval 2.4.6 Colour matching in plastics production 2.4.7 Criminal profiling 2.4.8 Identifying freshwater invertebrates 2.4.9 Interpreting electronic nose data 2.4.10 Electronics manufacturing diagnosis

2.5 CBR and fuzzy logic based weather prediction systems

2.5.1 CBR weather prediction systems 2.5.2 Fuzzy weather prediction systems

3. System for Fuzzy k-Nearest Neighbors Based Weather Prediction

3.1 Large database of airport weather observations
3.2 Fuzzy k-nn algorithm

3.2.1 Configure similarity-measuring function 3.2.2 Traverse case base to find k-nn 3.2.3 Make prediction based on weighted median of k-nn

3.3 Comparison to previous approaches

3.3.1 Fuzzy nearest prototype algorithm 3.3.2 Classic case-based reasoning

4. Experiments

4.1 Effect of varying attribute set
4.2 Effect of varying k
4.3 Effect of varying size of case base
4.4 Effect of varying fuzzy set membership function
4.5 System versus persistence

5. Conclusion

Additional References on Analog Forecasting in Meteorology

Appendix A: Sample Questionnaire for Knowledge Acquisition

Appendix B: A Worked-out Example of Fuzzy k-nn Algorithm for Prediction