Developed an end-to-end, large scale, Credit-Card Recommendation System
Devised a novel two-module clustering and rule-based model for generating recommendations from an unlabelled dataset, respecting user transaction patterns and maximising accrued reward points
Engineered a recommendation strategy that tackled problems such as cold-start and uncharacteristic spikes in transaction patterns
Proposed a data storage and processing solution for a scalable application
ML Project Assistant, BITS Pilani
Aug 2020 — Nov 2020
Conducted a literature survey of Delay Tolerant Networks followed by setting up a simulation of an open-sourced bus movement trace on the ONE simulator
Designed Decision Tree and Random Forest models to improve the performance of SaW(Spray-and-Wait) and Epidemic routing algorithms on the Delay Tolerant Network
Summer Research Intern, CSIR-CEERI Pilani
May 2020 — Jul 2020
Worked on the Capsule Network architecture for handwritten digit recognition
After a thorough literature review, implemented the Capsule Network architecture on the classic MNIST and the indigenous Kannada-MNIST datasets
Improved upon the state-of-the-art performance of Capsule Networks on the Kannada-MNIST dataset