Bank Card Usage Prediction Exploiting Geolocation Information

TL;DR

This paper presents a machine learning approach using gradient boosted decision trees and geolocation features to predict bank card usage, achieving top performance in a competitive challenge.

Contribution

The paper introduces a novel combination of hyperparameter tuning and geolocation-based features for bank card usage prediction.

Findings

Achieved first place in the first challenge task.

Secured fourth place in the second task.

Demonstrated effectiveness of geolocation features in prediction accuracy.

Abstract

We describe the solution of team ISMLL for the ECML-PKDD 2016 Discovery Challenge on Bank Card Usage for both tasks. Our solution is based on three pillars. Gradient boosted decision trees as a strong regression and classification model, an intensive search for good hyperparameter configurations and strong features that exploit geolocation information. This approach achieved the best performance on the public leaderboard for the first task and a decent fourth position for the second task.