Data Scientist
Abu Dhabi Islamic Bank
Total des années d'expérience :10 years, 6 Mois
Fragma Data Systems
Key Responsibilities:
•Working with global and major European banks.
•Involved in Campaign Analytics and Customer Targeting using Predictive Modeling and Advanced Analytics
•Developed SAS and Python based logistic regression models for cost effective targeting of customers in various promotional and cross sell campaigns.
•Validated the regression models using data from different time periods and different samples.
•Developed fully automated SAS macros for outlier treatment, missing value treatment, VIF based variable reduction.
Projects Undertaken:
•Finding Potential Credit Card Customers
•Predicted prospective buyers(Predictive Modeling)
•Differentiated the key characteristics between buyers and non-buyers.
•Focussed on predicting prospective buyers from existing customers
•Outlier treatment, Rank ordering(Scoring), KS Statistics and Variable reduction techniques on the basis of Clustering(PROC Varclus), WOE and IV.
•Built Logistic Regression model and studied the ROC curve and other generated parameters such as Concordant/Discordant, Accuracy, Precise, Recall and F-Measure, Betas etc.
•The final delivery of the project results showed tremendous predicting ability. During its first phase of implementation, the success rate was of the roof. The client almost doubled their existing profit by targeting the correct targets as suggested by the Predictive Model
Worked on consolidating data on dashboards, as well as creating statistical analysis comparing year to date, previous year to date and monthly models.
•Creating user forms in VBA and provided it to the manager so that he can view the reports of the team with click of a Button
Created Dashboards and monthly reports using MS Excel showing business revenue and scope of requirements from the clients.