An individual assignment for DTU course 02445, Project in Statistical Evaluation of Artificial Intelligence and Data.
The notebook predicts self-reported frustration scores from heart-rate features. It compares ridge regression and random forest regression with a baseline, uses subject-level cross-validation, and examines per-subject errors with statistical tests.
- Task 2 notebook — data exploration, modelling, evaluation, and statistical comparison.
The course-supplied HR_data.csv is not included in this public repository because it contains human-participant measurements. The notebook expects that file beside Task 2.ipynb, with the columns Individual, Frustrated, and heart-rate features beginning with HR_.
You can read the saved notebook on GitHub without the dataset. To rerun it, use a Python/Jupyter environment with pandas, NumPy, scikit-learn, SciPy, Matplotlib, and Seaborn, and provide a copy of the dataset you are authorized to use. The notebook is a course submission rather than a packaged application; dependency versions are not locked here.
The notebook groups validation folds by participant so measurements from one participant are not split across training and validation. Its results should be read in the context of this course dataset and assignment.