Portfolio
Machine Learning Projects
Credit Risk Model - ML Approach
Developed a credit risk model that helps financial institutions to identify probable Credit defaulters and help to maximize the net returns on loan portfolios. Moreover, this model would help the institution in mapping its existing portfolio into clusters of "low," medium, " and "high” risk for continuous tracking /improvements. Used Supervised Machine learning techniques like Decision trees, SVM Random Forest to build model code which predicts the Default customer and tie them to Cost matrix to optimize the Net Revenue function and mitigate risk for the Financial Portfolio (Mortgages)
Amazon Ecommerce – Case Study
Developed a customer segmentation model based on RFM (Recency, Frequency & Monetary) metrics and used Hierarchical clustering. Used Elbow Plot to decide the number of clusters and descriptive cluster analysis to segment customers based on their buying behaviour and patterns to help in developing marketing strategies (Increase engagement, Personalized Promotions, Incentives) according to the needs of each segment.
Pricing Analytics for Telecom Customer
Understood the consumer behaviour and their willingness to pay for different Service utilities (TV, Internet, Mobile) in marketplace and utilize this information to build a Linear model using Excel solver to optimize the Price points for individual services that were to be launched in the market. Highlighted the Power of bundling (2 or more services together) and the opportunity to further optimized revenue with a market surplus from existing clients.
Credit Risk Model - ML Approach
Developed a credit risk model that helps financial institutions to identify probable Credit defaulters and help to maximize the net returns on loan portfolios. Moreover, this model would help the institution in mapping its existing portfolio into clusters of “low,” medium, ” and “high” risk for continuous tracking /improvements. Used Supervised Machine learning techniques like Decision trees, SVM Random Forest to build model code which predicts the Default customer and tie them to Cost matrix to optimize the Net Revenue function and mitigate risk for the Financial Portfolio (Mortgages)
Amazon Ecommerce – Case Study
Developed a customer segmentation model based on RFM (Recency, Frequency & Monetary) metrics and used Hierarchical clustering. Used Elbow Plot to decide the number of clusters and descriptive cluster analysis to segment customers based on their buying behavior and patterns to help in developing marketing strategies (Increase engagement, Personalized Promotions, Incentives) according to the needs of each segment.
Pricing Analytics for Telecom Customer
Understood the consumer behaviour and their willingness to pay for different Service utilities (TV, Internet, Mobile) in marketplace and utilize this information to build a Linear model using Excel solver to optimize the Price points for individual services that were to be launched in the market. Highlighted the Power of bundling (2 or more services together) and the opportunity to further optimize revenue with a market surplus from existing clients.
Chronic Disease Patient Healthcare - Mobile App Salus
