Project Showcase
Highlight of my featured works
☀️🔋📊 Solar Panel Energy Prediction
A deep learning regression project predicting energy usage from solar panels. Includes EDA, statistical preprocessing, and neural network training with hyperparameter tuning.
🎯 What I Learned
- Performed statistical testing and preprocessing (VIF, Shapiro)
- Applied dense neural networks for regression
- Improved accuracy with Keras Tuner hyperparameter search
🛠 Tech Stack
- Python, Pandas, NumPy
- Matplotlib, Seaborn, Plotly
- Statsmodels, Scipy
- TensorFlow / Keras, Keras Tuner
📌 Evaluation / Next Step
The tuned neural network achieved lower MSE/MAE and higher R² compared to the baseline model. However, extreme weather variations are not fully captured, which can reduce performance in real-world deployment. Next steps include adding weather forecast features and testing sequential models like LSTM or GRU.