Lab exercises from the Uczenie Maszynowe (Machine Learning) course at AGH University of Kraków.
Thirteen lab folders covering the classic ML pipeline plus the first steps into deep learning with Keras / TensorFlow.
| # | Topic |
|---|---|
| lab01 | End-to-end ML workflow on the California housing dataset — exploration, preprocessing, train/test split, evaluation |
| lab02 | ML fundamentals follow-up |
| lab03 | Polynomial regression, regularization (ridge, lasso, elastic net) |
| lab04 | Classification on Iris and breast-cancer datasets; metrics; one-vs-rest |
| lab05 | Decision trees, ensembles, feature importance |
| lab06 | More tree-based methods, hyperparameter tuning |
| lab07 | Feature engineering and selection |
| lab08 | Model evaluation deep dive |
| lab09 | Neural networks with Keras / TensorFlow — Fashion-MNIST + regression |
| lab10 | Hyperparameter search with Keras Tuner (random + Bayesian) |
| lab11 | Convolutional Neural Networks |
| lab12 | Recurrent Neural Networks, time series |
| lab13 | Advanced deep learning topics |
cd labNN
jupyter notebook labNN.ipynb
# or run the script equivalent
python labNN.pypython · scikit-learn · pandas · numpy · tensorflow / keras · keras-tuner · matplotlib / seaborn.
This is coursework — implementations follow course tasks rather than original research. Trained model files (*.keras) and lecturer-provided PDFs are not committed.