🔆 Solar Energy Prediction Project

👤 Author: Rudranshu Pandey

---

📌 Objective: To build a machine learning model that predicts solar panel energy output based on 6 key features: temperature, humidity, wind speed, solar irradiance, panel efficiency, and system size.

---

🛠️ Tools & Libraries:

• Python 3.10
• XGBoost (regression)
• scikit-learn
• pandas, numpy
• Streamlit (for web UI)
• Joblib (for model saving)
• VS Code (dev environment)

---

📊 Feature Engineering:

• Selected 6 essential features to reduce complexity
• Removed irrelevant/complex fields from raw data
• Normalized missing values (if any) with .fillna()
• Ensured train and test used same column structure

---

🧠 ML Pipeline:

1. Load raw data using data_loader.py
2. Preprocess it using preprocessing.py (feature filtering)
3. Train an XGBoost model on selected features
4. Save the model using Joblib
5. Predict test set efficiency/output and generate submission.csv

---

🚀 Streamlit Web App (app.py):

• Allows manual input for real-time simulation
• Supports batch prediction from uploaded CSVs
• Displays predictions, alerts, and download options

---

📦 How to Run:

1. pip install -r requirements.txt
2. python main.py → to train and generate submission
3. streamlit run app.py → to launch the web interface

---

✅ Deliverables:

• Trained model (models/solar_model.pkl)
• Prediction output (submission.csv)
• Clean codebase with modular design