The objective of this work “Biclustering Analysis on Class Discovery from Gene Expression Data”s, is to analysis biclustering techniques of data mining to create new algorithm and to improve the existing algorithm in order to apply them to some of the issues in Gene Expression Data such as Class Discovery to minimize the complexities involved in providing solutions to them. In this work an algorithm is proposed Enhanced Coupled Two-Way Clustering Algorithm (ECTWCA) for Class Discovery from Gene Expression Data. Biclustering approaches adopted in this work identifies coherent patterns known as scaling and shifting patterns from high dimensional Gene Expression Data.