Defining, Predicting, and Preventing Disengaged Users in FinTech

Designing and developing machine learning models and AB testing experiments to reduce disengaged users in FinTech

Part I:

1. Defining engagement
2. Predicting engagement (data pipeline, feature extraction, and model training using pySpark and Scikit-Learn)
3. Model evaluation

Part II:
4. Designing intervention (contextual recommender system using ALS)
5. Designing A/B Testing
6. Evaluating and interpreting the results

Article published in: https://towardsdatascience.com/defining-predicting-and-preventing-disengaged-users-in-fintech-30dcb3bc0460

Singapore, May 2021