Simon's ML Portfolio

Machine-learning projects and study notebooks, published as a website with Quarto and GitHub Pages.

Live site: https://simoncan10.github.io/ML/

• projects/ — self-contained ML projects
  • home-credit-default-risk/ — Kaggle Home Credit competition: EDA, scikit-learn pipelines, feature engineering, and random-forest modelling. Notebooks in notebooks/, shared code in src/.
  • nlp-assignment/ — NLP coursework notebook.
  • natural-gas-prices/ — monthly natural gas price prediction: linear trend plus sinusoidal seasonality (scikit-learn).
• learning/ — study notebooks and exercises
  • hands-on-ml/ — worked chapters from Aurelien Geron's Hands-On Machine Learning.
• resources/ — notes and useful links (e.g. dataset sources).
• binder/ — environment pin for running the notebooks on Binder.

The Home Credit notebooks download the competition data from Kaggle on first run. Provide Kaggle credentials via the environment variables KAGGLE_USERNAME / KAGGLE_KEY, or a ~/.kaggle/kaggle.json file. See projects/home-credit-default-risk/src/data_loader.py for details.

Requires Quarto:

quarto render