Anjan Kumar Krishnan

Leveraging customer data to help estimate value from retaining customers, maximizing return on marketing expenditure through targeting and predictive modeling

• Which subscription model has higher customer lifetime value?




• How can customers' past behaviour inform a mobile app push messaging service?




• Next apparel category a customer is likely to purchase




• Which customers would respond best to a software upsell?




• How to automate a Data Quality Report (DQR) on tabular data?

Problems in forecasting and pricing analytics are explored

• Forecasting : How many cartons of eggs?




• How to price a golf ball product?

When random variables are aggregated to form a new performance metric (eg. total business profit), the aggregated random variable's distribution doesn't conform to a known probabality distribution. In such cases simulation modeling has been employed to calculate such random variables and aid business decision-making

• Insurance policy evaluation: Ontario gateway




• Discrete Event Simulation: Mine Site Productivity - Haulers in motion

Unsupervised and supervised machine learning algorithms are used o detect anomalies and fradulent transactions

• Which NYC property tax listings are anomalous?

Credit risk and underwriting models are created using statistical modeling (regressions) and machine learning techniques

• How likely is a credit card user likely to default on payment?