Bank of Jiangsu to build a smart credit card lab with Tencent Security

On April 17, Bank of Jiangsu and Tencent announced cooperation to explore so called Federated Learning. The two sides will jointly build an intelligent credit card management joint laboratory, to carry out in-depth cooperation around Federated Learning, promote the combination of AI technology and credit risk control, and open the door to intelligent credit card management.

On April 17, Bank of Jiangsu and Tencent announced cooperation to explore so called Federated Learning. The two sides will jointly build an intelligent credit card management joint laboratory, to carry out in-depth cooperation around Federated Learning, promote the combination of AI technology and credit risk control, and open the door to intelligent credit card management.

According to reports, federated learning is a new basic technology of artificial intelligence, which can effectively solve the contradiction between data privacy and the sharing of feature variables in cooperation, and ensure the information security in the exchange of feature variables in the cooperation of two or more parties.

Prior to this, Bank of Jiangsu and Tencent Security jointly developed and deployed intelligent credit card operations based on federated learning technology, and made technological breakthroughs in the large-scale access and profitability of credit cards on the premise of ensuring the information security of both sides. The system was launched on March 6, 2020.

The federated learning modeling takes only 3 days to deploy and the remote model training is completed in a few hours. Under comparison, the modeling efficiency has been greatly improved, and the ability to identify the fraud risk of Internet users is stronger and the coverage is wider.

Disclaimer: This is an article created by Michael Liang for Chinasdg.org. You can find the original article here: https://chinasdg.org/2020/04/19/bank-of-jiangsu-to-build-a-smart-credit-card-lab-with-tencent-security/.

Leave a Reply

Please Login to Comment