This project uses a deep learning approach with Bidirectional LSTM to detect spam emails from text data. It includes preprocessing, tokenization, model training with callbacks, and comprehensive evaluation.
- Tokenization & Padding
- Bidirectional LSTM with Dropout
- Model Checkpointing & Early Stopping
- Confusion Matrix & Classification Report
- Automatic Best Model Saving (.keras)
- Accuracy: 99%
- Python, TensorFlow, Keras, Scikit-learn, Pandas, Matplotlib, Seaborn
pip install -r requirements.txt
python spam_detection.ipynbYou can also access my Kaggle Notebook here >> https://www.kaggle.com/code/aryadev25/spam-email-detection