AgriRecoNet: Agricultural Recommendation Network 🌾

A clean, reproducible benchmark for the classic Crop Recommendation problem. This project implements and evaluates multiple machine learning baselines and a deep learning model to predict the optimal crop based on environmental and soil features.

---

✨ Highlights

This repository provides an end-to-end, ready-to-run solution for crop recommendation.

• Comprehensive EDA: Includes visualizations like a correlation heatmap to understand feature relationships.
• Multiple Baselines: Compares classic ML models: Logistic Regression, KNN, SVM, Decision Tree, Random Forest, and XGBoost.
• Deep Learning Model: Features a custom Artificial Neural Network (ANN) built with TensorFlow/Keras.
• Thorough Evaluation: Assesses models using accuracy, F1-score, confusion matrices, and ROC/AUC curves.
• Reproducible Environment: Comes with a requirements.txt file for easy setup.

---

📦 Dataset

The model is trained on the Crop Recommendation Dataset, which contains the following features:

• Soil Nutrients: Nitrogen (N), Phosphorous (P), Potassium (K)
• Soil Condition: pH value
• Environmental Factors: Temperature, Humidity, Rainfall

The target variable is the recommended crop label (e.g., Rice, Maize, Chickpea, etc.).

To Use: The notebook is configured to load Crop_recommendation.csv from a data/ directory.

1. Create a folder named data in the root of the project.
2. Download the dataset from Kaggle and place it inside the data/ folder.
3. Alternatively, update the path in the "Data Load" cell of the notebook to your specific file location.

---

🛠️ Setup & Run

Get up and running in a few simple steps.

1. Clone the repository:

   git clone [https://github.com/meanderinghuman/AgriRecoNet.git](https://github.com/meanderinghuman/AgriRecoNet.git)
   cd AgriRecoNet

2. Create and activate a virtual environment:

   # Create the environment
   python -m venv .venv
   
   # Activate it (Linux/Mac)
   source .venv/bin/activate
   
   # Activate it (Windows)
   .\.venv\Scripts\activate

3. Install the required packages:

   pip install -r requirements.txt

4. Launch the notebook: Open the AgriRecoNet.ipynb notebook in VS Code or Jupyter and run the cells.

   • In VS Code: Ensure you have the Python and Jupyter extensions installed. Select the .venv interpreter from the command palette (Ctrl+Shift+P).

---

🧪 Models Included

• Logistic Regression
• K-Nearest Neighbors (KNN)
• Support Vector Machine (SVM)
• Decision Tree
• Random Forest
• XGBoost
• Artificial Neural Network (ANN)

---

📈 Results & Evaluation

The models were evaluated based on their accuracy on the test set.

Model	Accuracy (%)
Naive Bayes	99.64
Random Forest	99.45
XGBoost	99.45
Gradient Boosting	99.09
Decision Tree	98.73
Artificial Neural Network (ANN)	98.18
Support Vector Machine (SVM)	98.00
Logistic Regression	97.27
K-Nearest Neighbors (KNN)	95.82

Evaluation is performed using:

• Accuracy Score
• F1 Score (Macro/Weighted)
• Confusion Matrix
• ROC / AUC Curves (One-vs-Rest for multi-class classification)

---

🗂️ Repository Structure

AgriRecoNet/
├── AgriRecoNet.ipynb
├── requirements.txt
├── .gitignore
└── README.md

---

🧭 Roadmap

Future enhancements planned for this project:

• Implement hyperparameter tuning sweeps (Optuna/Scikit-Optimize).
• Integrate cross-validation with stratification for more robust evaluation.
• Add model explainability using SHAP across all models.
• Create a script to export the best-performing model as a saved artifact for deployment.

---

📌 Notes

• TensorFlow Errors: If you encounter GPU errors with TensorFlow, try running on a CPU or ensure you have the correct CUDA/cuDNN versions installed for your hardware.
• XGBoost: This library is optional. If you face installation issues, you can comment out the XGBoost-related cells in the notebook.

---

🔑 Keywords

agriculture machine learning, crop recommendation, soil nutrients, nitrogen, phosphorous, potassium, pH, rainfall, temperature, humidity, XGBoost, Random Forest, SVM, KNN, ANN, TensorFlow, Keras, EDA, classification.