• Implemented Logistic Regression, Clustering and Random Forest techniques to predict frauds and 20% frauds are avoided with this model. Customers are further segregated on the basis of their fraudulent activities using Clustering technique that benefited client in coming up with new business problems.

• Implemented classification techniques for machine learning: Decision Tree, KNN and SVM technique (accuracy 95.7%, specificity 94.5%, sensitivity 97.1%). In conjunction with healthcare knowledge results help physician’s team to optimize treatment of patients so the cancer deaths are reduced.

Objective: Build a prediction model to identify the employees at the risk of attrition within the organization to undertake appropriate retention measures.

• Business Benefit: Implemented a Logistic Regression model, CART and Random Forest techniques to predict the employee attrition propensity in next 3 months. The results of model help HR Partners to retain at least 6 out of 10 high potential employees leading to reduction in the turnover and new hires cost.