Binary classification project predicting bank client churn using neural networks.
Check notebooks/churn.ipynb for models accuracy and additional performed tasks.
| Path | Description |
|---|---|
| data/ | Train / test CSVs (bank client features) |
| src/features.py | Preprocessing & feature engineering |
| src/my_neural_net.py | NumPy neural net built from scratch |
| src/NN_GridSearch.py | Keras residual net + hyperparameter grid search |
| src/DataSplit.py | Train/val/test split container |
| src/utils.py | Shared helpers |
| models/ | Saved Keras models |
| predictions/ | Output CSVs |
| notebooks/churn.ipynb | Exploratory analysis & experiments |
TensorFlow/Keras · scikit-learn · pandas · NumPy
uv sync
jupyter lab notebooks/churn.ipynb