Crop_prediction_based_on_predicted_rainfall_2024

This project involved prediction of rainfall level for 2024 year (location used: India) using model1 further this predicted rainfall levels for 2024 is fed into model 2 as test data for getting predicted crop levels for 2024.

Model Architecture

1. Model 1 (Random Forest Regressor)

   • Input: Historical rainfall data (2018-2023)
   • Output: Predicted rainfall levels for 2024
   • Features: Grid Search optimization for hyperparameters

2. Model 2 (Random Forest Classifier)

   • Input: Predicted 2024 rainfall data from Model 1
   • Output: Seasonal crop predictions for 2024
   • Features: Season-based segmentation

Data Sources

• Rainfall Data: Visual Crossing Weather Data Services
• Crop Data: Kaggle dataset (augmented with climate and fertilizer data)

Key Features

• Historical rainfall pattern analysis (2018-2023)
• Seasonal segmentation for Indian agricultural contexts
• Hyperparameter tuning using Grid Search
• Integration of real-time weather data

Model Performance

• Model 1 evaluated using:
  • Mean Absolute Error (MAE)
  • Mean Squared Error (MSE)
  • R-squared (R2) score