DELAY-CFIM：基于滑动窗口的高速数据流闭合频繁模式挖掘方法

DELAY-CFIM:基于滑动窗口的高速数据流闭 合频繁模式挖掘方法

DELAY-CFIM: A SLIDING WINDOW BASED METHOD ON MINING CLOSED FREQUENT ITEMSETS OVER HIGH-SPEED DATA STREAMS

摘 要

ABSTRACT

ACKNOWLEDGEMENTS

CONTENTS

CHAPTER 1 INTRODUCTION

1.1 Background

1.2 Literature review

1.2.1 Related work based on sliding window

1.2.2 Related work based on landmark window

1.2.3 Related work based on damped window

1.3 The main contents and organization of the thesis

CHAPTER 2 PRELIMINARY

2.1 Introduction

2.2 Closed frequent itemsets

2.2.1 Definition of closed frequent itemsets

2.2.2 Features of closed frequent itemsets

2.3 Sliding window

2.4 Conclusion

CHAPTER 3 ORDERLY TRANSACTION TREE AND CLOSED FREQUENT ITEMSET TREE

3.1 Introduction

3.2 The summary data structure - OTT

3.2.1 Definition of OTT

3.2.2 Construction and maintenance of OTT

3.3 Closed frequent itemset tree - CFIT

3.3.1 Definition of CFIT

3.3.2 Construction of CFIT

3.4 Conclusion

CHAPTER 4 ALGORITHM DELAY-CFIM

4.1 Introduction

4.2 Frequent itemset generation

4.2.1 Introduction of frequent itemset generation algorithm

4.2.2 Correctness proof for frequent itemset generation algorithm

4.3 Closure detection

4.3.1 Introduction of closure detection algorithm

4.3.2 Correctness proof for closure detection algorithm

4.4 Pruning

4.4.1 Pruning in frequent itemset generation

4.4.2 Pruning in closure detection

4.5 Conclusion

CHAPTER 5 EXPERIMENTAL RESULTS

5.1 Introduction

5.2 Datasets used in experiments

5.3 Experiments on sliding window

5.3.1 Experiments on different windowSize

5.3.2 Experiments on different min_sup

5.3.3 Experiments on different transaction length

5.3.4 Experiments on different query frequencies

5.3.5 Experiments on different datasets

5.4 Conclusion

CONCLUSION

REFERENCES

攻读硕士学位期间发表的学术论文

哈尔滨工业大学学位论文原创性声名及使用授权说明