🌾 Crop Yield Prediction

This repository contains a project for predicting crop yield using machine learning models (XGBoost, LSTM, Random Forest) based on climatic and economic variables.

---

📂 Project Structure

• 📓 Notebooks – Colab .ipynb files for data analysis, model training, and evaluation:

  • ApplesProjectEDA.ipynb – Exploratory Data Analysis and preprocessing.
  • ApplesProjectModels.ipynb – Model training and evaluation.

• 📈 Results – Performance comparisons and visualizations.

• 📁 Data – Dataset files (not included in the repository, see details below).

---

📌 Features

• ✅ Descriptive statistics (mean, median, standard deviation).
• ✅ Data distribution tests (Histogram, Q-Q Plot, Shapiro-Wilk, Pearson).
• ✅ Models: XGBoost, LSTM, Random Forest.
• ✅ Evaluation using RMSE vs. STD.

---

📈 Results Summary

Model	Test STD	RMSE	Zone / Approach
XGBoost	2.497	2.244	Individual zone – Yonatan
XGBoost + Econ	2.071	1.364	Combined zones + economic data
LSTM	2.497	2.316	Individual zone – Yonatan
Random Forest	2.532	2.271	Combined zones + economic data

• XGBoost with combined features outperformed other models.
• RMSE significantly lower than STD indicates high prediction accuracy.
• LSTM and Random Forest underperformed in this specific scenario.

---

🚀 How to Use

1️⃣ Clone the repository

git clone https://github.com/Dangutman98/Apple-Yield-Climate-Research.git
cd Apple-Yield-Climate-Research

2️⃣ Open the Notebooks in Google Colab

• Open ApplesProjectEDA.ipynb and run it step-by-step for EDA & preprocessing.
• Then run ApplesProjectModels.ipynb for model training and evaluation.

ℹ️ Note: The datasets are loaded directly from the authors' Google Drive during notebook execution. You do not need to upload any files manually. Simply run the notebook cells as-is.

3️⃣ Recommended Run Order

1. Run ApplesProjectEDA.ipynb (mandatory for data preparation).
2. Run ApplesProjectModels.ipynb (after EDA is completed).

---

👨‍💻 Authors

• Dan Gutman
• Tal Krispin
• Shahar Ben Laiche

Collaborators

• Volcani Institute – Israel's national powerhouse of agricultural innovation.

---

⚠️ Data Notice

• The dataset files are not included in this repository due to size constraints.
• The notebooks are pre-configured to fetch the required files from a private Google Drive. Ensure you have access via the shared links or modify paths accordingly.