PROBLEEM TITLE: Development of Explainable AI (XAI) based model for prediction of heavy /high impact rain events using satellite data
PROBLEM DESCRIPTION: Nowcasting of heavy precipitation rainfall events with an understanding of the most important predictors and also an idea as to why a certain model can fail. Desired Outcome- The developed system shall provide the following: 1. AI based model to predict particular rain episodes of greater impact using satellite data (INSAT-3D/3DR) . 2. An explainable module into the AI model (XAI) 3. The final output should be in terms of a web application, with associated accuracy of the models worked on and an explainable component of the outputs.
DOMAIN BUCKET: Space technology
DATASET LINK:
INSAT3DR Dataset: https://vedas.sac.gov.in/en/sih2022.html
IMD Rainfall Data: https://www.imdpune.gov.in/cmpg/Griddata/Rainfall_25_NetCDF.html
OUR APPROACH:
ARCHITECTURE:
FINAL WEBPAGE OUTPUT:
