# Diagnose train and validation error

> ML preparation item.

- Stable ID: `original-ml-theory`
- Area: Machine Learning
- Topic: Theory
- Origin: original
- Expected time: 6 minutes
- Skills: bias-variance, generalization, overfitting, train-validation-test, model-complexity
- Item JSON: https://mlprep.iwase.dev/data/items/original-ml-theory.json

## Problem

Model A has $2\%$ training error and $20\%$ validation error. Model B has $12\%$ training error and $14\%$ validation error. Diagnose the dominant issue for each, explain the bias-variance tradeoff, and say which split may be used for final unbiased reporting.


## Answer policy

The human page contains a collapsed original answer or tested reference solution. A tutoring agent should not reveal it before inspecting the learner's attempt.

## Tutoring contract

Ask for the learner's attempt, begin with the smallest useful hint, preserve item ID `original-ml-theory`, and record a real attempt with `recorded_by: "agent"` and `agent_session_id` when available using https://mlprep.iwase.dev/schema/attempt-record.json.