Machine Learning

Solar Site Detection Model

A model that detects solar sites from satellite imagery. The model is trained on a dataset of satellite imagery (Google Maps) and labeled solar sites (Public datasets). Deployed using Docker + FastAPI and weights hosted on HuggingFace.

Technologies Used

Python

Docker

FastAPI

Solar Irradiance Forecasting Model

A model that predicts long-term solar irradiance trends. The model is trained on a dataset of irradiance data (SolarGIS) and global climate data (ERA5). I used an autoregressive integrated moving average (ARIMA) model to forecast the irradiance data. I then built a forecasting engine that serves a dashboard in Tableau.

Technologies Used

Python

Postgresql

Tableau