Learning by building.
A selection of project explorations across machine learning, computer vision, forecasting, and sports data. Repositories are linked on each card.
MNIST Handwritten Digit Prediction
Trained a PyTorch ANN on MNIST to classify handwritten digits from 0 to 9; added a Streamlit image-upload interface for predictions.
Plant Disease Detection
Built and trained a PyTorch CNN on PlantVillage leaf images to classify 15 disease classes; added a Streamlit image-upload interface.
Electricity Demand Forecasting
Compared RNN, LSTM, and GRU models to forecast next-hour household electricity use from the previous 24 hourly Global_active_power readings.
IPL Data Analysis
Cleaned and integrated IPL datasets for season, match, team, and player analysis, then performed exploratory analysis and created visualizations.
Project descriptions stay focused on the work areas; no performance metrics or results are claimed here.