A Proposal Method for Missing Value Analysis: Cluster Analysis Approach

Imputing values to missing cases is a subject that frequently met in the fields of Machine Learning and Data Mining, require researchers study it. It known many computer-based analysis algorithms operate under assumption there no case. The lack sufficient search case by able negatively affect performance results. In this study, it was studied with data set consisting 52 variables order measure Corporate Sustainability district municipalities Istanbul. Little’s MCAR applied on 17 containing case, determined were MCAR, namely completely at random. And then Clustering Analysis 35 not imputations made based clusters formed. observed cluster labels municipalities, whose clustering results obtained didn’t contain following imputation same, change. change indicates formed doesn’t draw away from main data, structure get disrupted. Consequently, can be said effective terms imputing more representative