These are some of the projects I have worked on in the past. For more visit my Github Profile
Trained a Isolation Forest anomaly detection algorithm which gave 96.5% accuracy in predicting fraudulent transactions on test set using Sklearn in Python. The dataset was a higly skewed dataset of 284,807 transactions in which only 492 transcations were fradulent (0.172%).
PyReData is a Python module which allows Python users to share their analysis and visualizations done using Python as HTML web pages. The module currently provides support for pandas , matplotlib and seaborn. I am currently in the process of re-writing the module.
Analyzed Amazon unlocked Mobile Reviews dataset of 400,000 reviews for relationship between brands,pricing and ratings. Developed visualizations using Seaborn and matplotlib in Python. Used NLTK for processing text reviews. Working on more in-depth analysis of dataset.
Worked with a team of three to develop a solution for managing disasters effectively. Solved transparency problem using Blockchain and remote connectivity using Peer-To-Peer based Mobile application. As part of the project I Developed a web Application for Relief Centers using NodeJs , Express , Bootstrap 4 and MongoDB
A Library for Private, Secure, Multi-Owner Deep Learning - Currently Pre Alpha. I am contributor to the open source library written in Python. Previously I worked on code documentation and refractoring codebase with type annotations. I worked on development of Polynomial tensor for non-linear multi-party computation operations. Further , I wrote an implementation of PATE (Private Aggregate of Teacher Ensembles) in PyTorch.
I have taken up some freelance web development work. These are the past projects I have completed.