Spam Email Classifier (GRU + LSTM)
This model predicts whether an email is Spam or Non-Spam.
Model
- Architecture: Embedding + GRU (150)+GRU(125) + Dense
- Trained on: Labeled spam and non-spam emails dataset
Input
- Raw text email (string)
- Convert to integer sequences using the same tokenizer
- Pad sequences to
max_len
NOTE:
This model work more finely on long email (100-1000 word).
Output
- Probability between 0–1
- Closer to 1 → Spam
- Closer to 0 → Non-Spam
Social Media Link
