Personalized marketing using machine learning at MassMutual

MassMutual defines individuals within 3 distinct behavioral segments in order to determine the appropriate marketing material to mail them.

Customers are divided into one of the three segments of Assured, Aspirational, or Apprehensive utilizing a data product from third-party vendors along with internal data assets. The dataset has a high rate of incomplete matches to our internal data sources, so we reproduce the segments for all unmatched individuals. The models were learned from matched data and applied to unmatched data. To address some class imbalance issues, SMOTE tool was applied. The best model was based on Gradient Boosting algorithm via the xgboost package.

Event Schedule:

• Doors at 6:15 pm (there will be someone downstairs checking you in)
• Talk begins promptly at 7 pm with Q&A following
• Networking & Drinks!

Food & beverages will be available!

-------- Sponsored by Comet.ml ---------
Comet.ml is doing for ML what Github did for software development. We allow data science teams to automagically track their datasets, code changes, experimentation history and production models creating efficiency, transparency, and reproducibility.

See how Comet can help your data science team - http://bit.ly/2W0bkBc