Diff for "ClusteringAlgorithms"

Back to ComputerTerms, InformationRetrieval

Cluster analysis is a statistical technique used to generate a category structure.The groups which are formed should have a high degree of association between members of hte same group and a low degree between members of different groups.

Similarity Measures:

               2C
 S         = -------
  (Di,dj)     A + B 

   Where C is the number of terms that Di and Dj have in common, 
   and A and B are the number of termsin Di and Dj

Similarity Matrix calculates a similarity measure between document x and y

  | S21                 |
  | S31  S32            |
  |  ...                |
  | SN1  SN2 ...SN(N-1) |

Clustering Methods

Nonhierarchical methods (partitioning methods) divide the data set of N objects into M clusters, where no overlap is allowed. Since discovering the optimal grouping of N objects into M sets is Combinations(n|M)=N!/((N-M)!M!) is impossible for large sets, so we use heuristics. Computational requirements are then usually O(NM)

Hierarchical methods produce a nested data set in which pairs of items or clusters are successively linked until every item in the data set is connected. The agglomerative method performs in N - 1 pairwise joins beginning from an unclustered dataset.

Back to ComputerTerms, InformationRetrieval