Optimized kmeans clustering with intelligent initial centroid selection

Optimized k-means clustering with intelligent initial centroid selection for web search using URL and tag contents WIMS ‘ 11 Proceedings of the International Conference on Web Intelligence, Mining and Semantics, Article No. 65, ACM, 2011 S. Poomagal Research Scholar, PSG College of Technology, Coimbatore, Tamilnadu, India T. Hamsapriya PSG College of Technology, Coimbatore, Tamilnadu, India 1 2013 / 05 / 07 haseshun

Index 1. Introduction 2. K-means Clustering Algorithm 3. Proposed Method 4. Experimental Result 5. Conclusion 2

Index 1. Introduction 2. K-means Clustering Algorithm 3. Proposed Method 4. Experimental Result 5. Conclusion 7

2. K-means Clustering Algorithm • 例 5. 変化がなくなるまで繰 り返す。 15

Index 1. Introduction 2. K-means Clustering Algorithm 3. Proposed Method 4. Experimental Result 5. Conclusion 17

3. Proposed Method 4. クラスタリングアルゴリズムの適用 • 例 1. TFIDF行列からscale factor(SF)を計算する。 TFIDF行列 D 1 T 2 T 3 T 4 T 5 4 7 9 2 1 0 2 D 2 6 D 3 3 1 0 1 4 2 5 3 6 1 0 1 3 1 5 1 8 K = 2と仮定する。 それぞれの用語のSF T 1 : MAX(T 1) / 2 = 8 / 2 = 4 T 2 : MAX(T 2) / 2 = 14 / 2 = 7 T 3 : MAX(T 3) / 2 = 36 / 2 = 18 T 4 : MAX(T 4) / 2 = 20 / 2 = 10 T 5 : MAX(T 5) / 2 = 18 / 2 = 9 30

Index 1. Introduction 2. K-means Clustering Algorithm 3. Proposed Method 4. Experimental Result 5. Conclusion 34

Index 1. Introduction 2. K-means Clustering Algorithm 3. Proposed Method 4. Experimental Result 5. Conclusion 39