One ML tutorial. Re-explained for your field, without losing the fundamentals.
Two short Jupyter notebooks teach linear regression and a tiny neural network. A VS Code Copilot Chat skill reads a learner's background and rewrites only the explanation cells — analogies, examples, exercises, and copy-pasteable “go deeper with an LLM” prompts — while a glossary-lock keeps every canonical ML term identical across versions.
Below are five pre-rendered adaptations. Pick the one closest to your background, or clone the repo and run the skill on your own background.md.
Clone the repo, open it in VS Code, then in Copilot Chat invoke the /adapt-notebook skill with a one-line description of yourself — the skill will interview you, write your background.md, and adapt both notebooks to your field. The more detail you give about your role, day-to-day work, and what you want to learn, the better the analogies and examples will land.
Example — one-liner that triggers the interview:
/adapt-notebook I'm a marine biologist
Example — richer one-shot description (skips most of the interview):
/adapt-notebook I'm a postdoc marine biologist studying coral reef recovery. I spend most of my week running underwater transects, fitting growth curves to bleaching-recovery data in R, and writing grant reports. I'm comfortable with stats and basic Python. Analogies that would land: reef-fish abundance counts, dose-response to temperature stress, logistic growth of coral cover. Goal: be able to fit my own ML models to next season's transect data.
Example — reuse a saved background:
/adapt-notebook use examples/marine_biologist/background.md