CochranQTest<T>

Models & Types · AiDotNet.Evaluation.Statistics

Cochran’s Q test for comparing multiple classifiers on the same dataset.

Implements: IClassifierComparisonTest<T>

For Beginners

Cochran’s Q test is an extension of McNemar’s test:

  • Compares 3+ classifiers on the same binary classification task
  • Tests if all classifiers have the same error rate
  • Non-parametric test for matched samples
  • Uses chi-square distribution for p-value

How It Works

When to use:

  • Comparing multiple classifiers on the same test set
  • Binary classification only (use Friedman for continuous scores)
  • When samples are matched (same data across all classifiers)

Constructors

Constructor Summary
CochranQTest()

Properties

Property Summary
bool IsNonParametric { get; }
bool IsPaired { get; }
string Description { get; }
string Name { get; }

Methods

Method Summary
StatisticalTestResult<T> Test(T[][] predictions, T[] actuals) Tests if multiple classifiers have significantly different error rates.