Loto6 Deep Learning Prediction

A deep learning experiment for predicting Japan's Loto6 lottery numbers using BiLSTM and Monte Carlo Dropout.
No guarantee of winning. Use responsibly.

---

📌 Features

• Automatic fetch of the latest Loto6 results
• 4-layer BiLSTM with L2 regularization + Monte Carlo Dropout
• Regression output with uncertainty estimation (mean ± std)
• 1 main prediction + up to 3 Monte Carlo-based subsets

---

🚀 How to Run (Google Colab Recommended)

1. Click the Open in Colab badge above to open the notebook Loto6_DeepLearning.ipynb located at the root of the loto6-deeplearning repository.
2. Run Runtime → Run all from the top of the notebook.
3. Check the learning curves and the predicted numbers.

---

⚠️ Disclaimer

• This notebook is for research and educational purposes only
• No guarantee of winnings, use at your own risk
• Please gamble responsibly

---

🏷️ Requirements (Colab default)

• numpy
• pandas
• matplotlib
• tensorflow >= 2.12
• scikit-learn