Introduction: In the field of Medical data analysis Data Mining plays an influential role. We should be capable of extracting fruit-bearing information from wealthy medical data. Extraction of effective information from wealthy medical data and making the valuable decision for predicting the diseases increasingly becomes necessary. Missing data or incomplete data pose a great problem in analysis. There is a good number of traditional methods available for taking care of data cleaning. Objective: In this paper, we have attempted to throw light on various methods/tools existing for data cleaning. In our work we have imputed the missing values using different machine learning techniques and also have performed a comparative study of different machine learning techniques used. Methods: A total of five hundred records of liver cirrhosis patients is collected. Two tasks have been carried out here, one is imputing missing values and the other is finding classification accuracy. The data set with no missing values for the predictor variables are used to generate the regression equation. In Random forest multiple decision trees were built and then these trees are merged to get the more accurate and stable prediction. Results: We observed accuracy of class prediction before imputing the missing values and after imputing the missing value by using different algorithms. Conclusion: It is noticeable that the accuracy of class prediction is high when missing values are handled properly. Also, the efficiency of class prediction is very high when the random forest is used both for imputing the missing values as well as for predicting the class.