Competition was Classify This Movie by Genre Prediction of Movies based on their Poster

Histopathologic cancer analysis and predictive model.

This repo consists of notebooks I created to work upon the skills I learnt through several courses. All notebook has EDA, Model Building, Hyperparameter Tuning, Ensemble Models and Sampling Techniques.

Approaching (Almost) Any Machine Learning Problem

C++ OpenMP and CUDA implementation of parallel ensembling of global stiffness matrix.

An ensemble is based on the notion of combining models. While uplift modeling combines supervised modeling with A-B testing, which is a simple type of randomized experiment.

Metis project 4/7

Analysis of Insurance Liability Claim Amount for settlement

Deepfakes are synthetic media in which a person in an existing image or video is replaced with someone else's likeness. We make a model which is efficient in detecting one.

This is 35th place solution to shell ai hackathon hosted on HackerEarth platform.

Templated boilerplate for experiments in Active and Ensemble Learning.

Visually-Assisted Performance Evaluation of Metamodels in Stacking Ensemble Learning

Ensemble methods for improving extractive summarization of legal case judgements

An AutoRecSys Library built around LensKit. Performs automatic algorithm selection, hyperparameter optimzation and ensembling on LensKit models.

37th place solution for the 2020-2021 Kaggle competition "Rock, Paper, Scissors"

: Visualization & Prediction

An named-entity-recognition (NER) based anonymizer for archival documents metadata.

It is From Analytics Vidhya Hackathons, Sponsored by Club Mahindra. It is based on Regression Problem, Where Accuracy matters the most, It is measured by RMSE Score. Different Techniques such as Stacking, Ensembling, Boosting and Scientific Operations such box-cox Operations to reduce skewness of the data.

Decision Trees, Ensembling, Adaboost and Random Forest on Bitcoin Heist Ransomware Address Dataset