# Theory Practice

> ML preparation set.

- Stable ID: `theory-practice`
- Area: machine-learning
- Kind: practice
- Timebox: 25 minutes

Reason about bias, variance, overfitting, generalization, and evaluation splits.

## Instructions

1. Use observed train-validation patterns rather than vague model-size rules.

## Ordered items

1. [Diagnose train and validation error](https://mlprep.iwase.dev/item/original-ml-theory/?set=theory-practice) — `original-ml-theory` (6 min)
2. [L1 versus L2 regularization](https://mlprep.iwase.dev/item/original-ml-regularization/?set=theory-practice) — `original-ml-regularization` (6 min)
3. [Read a finite-class generalization bound](https://mlprep.iwase.dev/item/original-ml-finite-class-bound/?set=theory-practice) — `original-ml-finite-class-bound` (7 min)
4. [Generalization and overfitting knowledge check](https://mlprep.iwase.dev/item/google-mlcc-overfitting-quiz/?set=theory-practice) — `google-mlcc-overfitting-quiz` (10 min)

## Completion

Recommend a defensible response to one bias and one variance failure mode.