🌱 Crop Yield Predictor – Databricks Edition
Description: The Crop Yield Predictor is an AI-powered application built entirely inside Databricks to predict crop production. Users can input State, District, Season, Crop, and Area, and get real-time yield predictions in tonnes. The model is a RandomForest MLflow model trained on multi-crop data. All predictions are logged in a Delta table for historical analysis and insights.
⸻
⚡ Features • 🌾 Predict production for multiple crops • 🖥️ Interactive Databricks notebook UI using widgets • 🤖 Real-time prediction with MLflow model • 📊 Predictions logged in Delta table for analytics • 📈 Historical trends and crop insights • 🚀 Fully self-contained — no external deployment required
⸻
🛠️ Technologies Used • Databricks – Notebook environment & UI • Python (pandas, scikit-learn) – Data processing & ML • MLflow – Model logging & versioning • Delta Lake – Store predictions & track history
⸻
🎯 How It Works 1. User selects State, District, Season, Crop, Area via notebook widgets 2. Input is processed and aligned with model features 3. RandomForest model predicts expected production in tonnes 4. Predictions are displayed instantly in the notebook 5. All inputs & outputs are logged in a Delta table for tracking
⸻
🚀 Getting Started 1. Open the notebook in Databricks 2. Run the setup cell to import libraries 3. Select input parameters using the widgets 4. Run the prediction cell → view predicted yield 5. Explore logged predictions via SQL or Delta tables
⸻
📌 Use Case • Helps farmers estimate crop yield • Assists policymakers in agricultural planning • Enables researchers to analyze trends & crop patterns
⸻
💡 Future Enhancements • Suggest alternate crops if predicted yield is low • Add graphical insights for districts & seasons • Expand to more crops and regions
Log in or sign up for Devpost to join the conversation.