Narendra N
Lead Consultant, AI Research
Wipro Limited

Abstract—Data driven analysis using Machine Learning (ML) and Deep Learning (DL) has gained
a lot of popularity in recent years. The success of these techniques in solving tasks in the field of
image processing, computer vision and natural language processing has made these techniques
almost indispensable. These techniques are now being explored in the areas of Finance,
Medicine etc. Application in these areas pose a specific problem of privacy. Privacy in financial
domain might be the hesitancy of a client to share his/her financial information for data analysis.
In the medical domain, the patient records are bound by confidentiality. Data encryption could be
used to preserve the privacy of such data. However, can we do analysis on encrypted data?
Homomorphic Encryption (HE) might be an answer to this question. HE is a form of encryption
which allows us to do computation on encrypted data. It is a lattice based cryptographic
technique which is gaining popularity in the ML research community to see if it can be
incorporated to solve privacy preserving ML tasks.