Multilayer perceptron

Wisconsin Breast Cancer (Diagnostic), UCI: 569 tumours described by measurements of cell nuclei, labelled benign or malignant.

Solo

1. A Keras-like API: a CustomSequential model stacking Dense layers.
2. Forward pass and backpropagation, with binary cross-entropy, categorical cross-entropy and MSE losses.
3. SGD and Adam optimisers, and early stopping on validation loss.
4. A Colab notebook that trains and evaluates the model end to end.

Mirror the Keras API

I followed the structure of the official Keras documentation, with layers, losses and optimisers as separate classes behind one model object. It kept the code modular instead of a growing pile of functions.

97.4 % validation accuracy on one 80/20 split, also used for early stopping.

• The validation split also drives early stopping, and there is no separate test set, so these scores are optimistic.
• The logs show overfitting: training accuracy reaches 1.0 while validation loss rises from 0.20 to 0.73.
• One split of 569 samples: a different seed could move the score by a few points. Cross-validation would give a range.