Joint Electronic Teller Services Limited (JETCO) (Sponsor)
Although the potential benefits of sharing enterprises' data to machine learning technology are huge, the data privacy and compliance are stopping the realization. Federated learning is seen as the last mile of machine learning launch that can tackle the data privacy issue. The recent advancements in federated learning can resolve the privacy-preserving problem of consolidating results of distributed machine learning models in private data sets owned by different enterprises [1]. However, enterprises are reluctant to engage in this technique due to the lack of application integration support. This project aims to design and implement an effective application integration software platform that facilitates the deployment of a federated learning network that is compliant with data privacy regulations so that the new use cases of machine learning can be deployed. The objective of the formation of the software platform is to help enterprises to upgrade their applications, such as credit assessment and anti-money laundering, from using individual machine learning models to federated learning models. This platform will be beneficial for a wide variety of users that include financial regulatory authorities, organisations using AI technology, research communities, and the public. [Reference] [1] Yang, Qiang, et al. "Federated machine learning: Concept and applications." ACM Transactions on Intelligent Systems and Technology (TIST) 10.2 (2019): 1-19.