A selection of technical projects in machine learning, networks, automation, and cycling tools.
Five mini-projects covering cross-validation, regularization for linear regressions, imbalanced and incomplete data, feature selection, deep neural networks, and image-oriented model structures using scikit-learn and TensorFlow.
Weekly exercises on graph generation, graph measures, and network applications using Python with networkx and pathpy.
A command-line tool that monitors ASVZ classes, waits for registration to open, and completes the sign-up flow automatically with selenium, pandas, and requests.
A command-line tool to estimate anaerobic capacity and anaerobic threshold from cycling power data and visualize the fitted power curves.
A route-time estimation tool based on a physical cycling model that combines GPX routes, rider power profiles, gradients, and curves, then renders both command-line output and a PDF report with map and elevation profile.