Hybridization of different soft computing techniques is very useful in solving real world problems. Soft computing techniques like rough set theory and fuzzy sets are used to handle uncertain or vague information present in the real world data. Hybridizing the soft computing techniques yield better result than the stand alone techniques. Hybridizing these techniques serves different purpose like attribute subset selection, improving classification accuracy, prediction and many more. Data mining concepts are useful in extracting knowledge from the large data set. The usages of some of the soft computing concepts hybridized with data mining concepts used for classification and attribute reduction are also focused and the future work of each concept is highlighted. Comparative analysis made with the survey work is focused which focuses on scalability, robustness, efficiency and accuracy of different classifiers. © 2016, International Journal of Pharmacy and Technology. All Rights Reserved.