My Projects
Welcome to My Project Showcase!
Hello and thank you for visiting my project page! 🌟 I'm excited to share my journey, passion, and insights through the projects I've worked on. Here, you'll find a collection of my work in data science, machine learning, and data visualization. Each project represents a step forward in my quest to uncover meaningful patterns and tell compelling stories through data. I hope you find something here that inspires you!
Exoplanet habitability and earthlikeness prediction
What if the next habitable world is just a few light-years away?
This project harnessed machine learning to evaluate the potential habitability and Earth-likeness of exoplanets, using key astronomical data. By developing a model that achieved 85% accuracy through classification and regression techniques, and optimizing it with advanced feature engineering, I contributed to the search for planets with Earth-like qualities, potentially expanding our understanding of life beyond Earth.
Potato Disease Classifier
Can artificial intelligence help farmers protect their crops from devastating diseases?
In this project, I developed a deep learning model using convolutional neural networks (CNNs) to classify potato plant diseases from images with high accuracy. Leveraging data augmentation techniques, the model was further optimized for robustness. The deployment was accomplished using Docker, Flask, and FastAPI, providing an interactive dashboard that enables real-time disease detection and actionable insights for farmers and agricultural experts.
Spider-Verse Classifier
​This project involves developing a machine learning classifier to identify Spider-Man actors from the Spider-Verse series based on their images. Leveraging a custom dataset gathered through a Pinterest scraper, the classifier uses feature extraction techniques and traditional machine learning algorithms to differentiate between actors like Andrew Garfield, Tobey Maguire, and Tom Holland. The model's accuracy improves with larger datasets, showcasing its capability to adapt and refine predictions. The implementation integrates image preprocessing, feature extraction, and classification to deliver reliable results. This project exemplifies the application of machine learning in a creative context, merging pop culture with advanced data analysis techniques.
Smart Course Recommendations Leveraging NLP and Clustering Techniques
​Ever felt lost in a sea of online courses, unsure which path to take? This project aims to simplify that choice by developing a smart recommendation system using Natural Language Processing (NLP) and clustering algorithms. The system analyzes thousands of course descriptions and clusters similar courses, providing tailored recommendations based on user preferences. Achieving 85% recommendation accuracy, this solution enhances user engagement by 30% and offers personalized learning journeys for each individual.
PhonePe Pulse Data Visualization and exploration
​How do you make sense of millions of transactions happening across India? This project tackles the challenge by visualizing transaction data from the PhonePe Pulse repository. Using Python, Streamlit, and Plotly, the project creates an interactive dashboard that dynamically updates from a PostgreSQL database. The dashboard provides actionable insights into transaction trends, regional growth, and user behavior, empowering businesses and analysts to make data-driven decisions with ease.
YouTube Data Harvesting and Warehousing
Ever wondered how much data YouTube generates every second? This project dives deep into the massive world of YouTube data by developing a system to extract, transform, and warehouse video data for analysis. Utilizing Python, SQL, and data warehousing techniques, the project automates the harvesting of data, storing it efficiently in a structured database. The system enables rapid querying and visualization of key metrics like video popularity, viewer demographics, and channel performance, offering valuable insights for content creators, marketers, and analysts alike.
