Massachusetts Institute of Technology (MIT) researchers have developed a machine learning technique that can drastically reduce false positives in fraud-detecting technology used by consumer banks that incorrectly flag credit card sales as suspicious.
The new system utilizes an "automated feature engineering" approach that extracts more than 200 detailed features from each individual transaction, enabling better identification when a specific cardholder's spending habits deviate from the norm.
The researchers tested the program on a dataset of 1.8 million transactions from a large bank, and found the number of false positive predictions was 54% lower compared to traditional models. They estimate that this could have saved the bank about $220,000 in lost revenue.
MIT's Kalyan Veeramachaneni said, "We can say there's a direct connection between feature engineering and [reducing] false positives. That's the most impactful thing to improve accuracy of these machine learning models."
From MIT News
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA