Overview / Usage

Machine Learning is nowadays used in everything we see online, and why not, it learns our actions. But, applying machine learning to a problem which involves medical, financial, or some other type of sensitive data, not only requires accurate predictions but also careful attention to maintaining data privacy and security. The solution is secure machine learning, in which the cloud predicts the results based on a machine learning classifier, but without compromising our privacy. This allows a data owner to send their data in encrypted form to a cloud service that hosts the network. The encryption ensures that the data remains confidential since the cloud
does not have access to the keys needed to decrypt it.

Methodology / Approach

In this project, we implemented machine learning approaches on encrypted datasets, using Homomorphic Encryption. Our approach included only the testing stage till now. We assume that the cloud already has a model that was trained in some way, for example using a set of unencrypted data. We analyzed the mathematical challenges behind implementing such secure models and successfully designed secure classifiers, applying them to real-life problems like privacy- preserving disease diagnosis, and stock price prediction using encrypted stock data.