Satellite Image Segmentation using UNet and Transfer Learning

This project tackles semantic segmentation of high-resolution satellite imagery using UNet and VGG19-UNet to accurately classify six land covers including vegetation, water, buildings, lands and roads. Demonstrated a full end-to-end pipeline from data preprocessing to model evaluation.

Energy Demand Forecasting - Time Series Forecasting

This project creates a system to predict hourly electricity usage on a monthly basis using historical data. I explored and compared three different models: ARIMA, Random Forest, and LSTM, and fine-tuned them to improve accuracy. The best-performing model was then used to create a Streamlit web application, making the predictions easy for anyone to access and use.

Sales, Customers and Product Analysis - SQL and PowerBI

This project focuses on building a modern data warehouse to generate actionable business insights. I used SQL for the full data lifecycle, from cleaning and transforming raw data to creating a structured database. These validated insights were then presented through interactive Power BI dashboards to drive data-informed decisions.

Python

Data Visualisation

SQL

Statistics

Data Cleaning and Wrangling

Machine Learning

Deep Learning

Post Graduate Program in Data Science and Analytics

Data Analysis with Python - IBM

AI/ML for Geodata Analysis - IIRS, ISRO

Summer Analytics Course 2024, IIT Guwahati

Python for Data Science, AI and Development