Full Stack Projects
A short summary of my previous full stack projects.
INVENTORY MANAGEMENT - AWS
This project is a user-friendly inventory management developed with NodeJS, NextJS, PostgreSQL and deployed using AWS.
MOVIE MINDS
MovieMinds is a full-stack web application built with the MERN stack, enabling users to explore trending and upcoming movies, manage profiles, and connect with cinephiles. It features a machine learning-powered recommendation engine, integrated with TMDB API, for personalized movie suggestions and dynamic user experiences.
REACT-BOOKSTORE
The React Bookstore project is a modern web application developed with React.js, featuring a dynamic UI designed in Figma and implemented with React Router, hooks, and context. It includes a robust Java backend using Jakarta EE and MySQL for secure data storage and efficient RESTful API integration.
HOUSE PARTY WITH SPOTIFY
Spotify Room Music Controller is a web application that integrates with the Spotify API to enable collaborative music playback. Hosts can create rooms, customize settings, and control playback, while guests join with a unique code to vote on skipping songs, ensuring an interactive and engaging group music experience.
Data Science Projects
A short summary of my previous Data science projects.
Patient Readmission Analysis
Developed a machine learning model achieving 93.3% accuracy in predicting hospital readmissions by preprocessing and engineering features from over 91,000 records, while leveraging Random Forest, XGBoost, and SHAP analysis for actionable healthcare insights.
Amazon Review Analysis
Performed sentiment analysis on imbalanced Amazon reviews using NLP techniques, deep learning models like CNN and RNN, and Gensim embeddings, achieving a 94% ROC AUC score through advanced text preprocessing and hyperparameter tuning.
Reddit Data Analysis
Analyzed Reddit data using topic modeling (LDA), Named Entity Recognition (NER), and sentiment analysis, and developed predictive models with SVM and XGBoost to estimate post scores, leveraging features like topics, embeddings, and sentiment trends for actionable insights.
Twitter - Sentiment Analysis
Performed sentiment analysis on Twitter data using advanced preprocessing techniques and multiple embedding methods, including Bag of Words, TF-IDF, Word2Vec, GloVe, and Universal Sentence Encoder, combined with SVM classification to achieve accurate sentiment predictions.