K-MEANS is a partitioning clustering algorithm in data mining which is very useful technique to find the nearest clusters in data. K-MEANS is an unsupervised learning, which use for divide the data into K-clusters. BIRCH is a Balanced Iterative reduction and clustering Hierarchies’ algorithm. It is a hierarchical based clustering in which is used to divide a large data in small clusters. To improve the quality to large dataset in which some values are not present we used a combination of K-MEANS and BIRCH algorithm to solve this problem. In this paper, we discussed the data set in which we do not get the exact problem in data set and how we solved it. To solve this type of problem we use represent a combination of K-MENAS and BIRCH algorithm.