No free lunch for cross validation

Huaiyu Zhu, Richard Rohwer

    Research output: Contribution to journalArticlepeer-review


    It is known theoretically that an algorithm cannot be good for an arbitrary prior. We show that in practical terms this also applies to the technique of ``cross validation'', which has been widely regarded as defying this general rule. Numerical examples are analysed in detail. Their implications to researches on learning algorithms are discussed.
    Original languageEnglish
    Pages (from-to)1421-1426
    Number of pages6
    JournalNeural Computation
    Issue number7
    Publication statusPublished - 1 Oct 1996

    Bibliographical note

    Copyright of the Massachusetts Institute of Technology Press (MIT Press)


    • learning algorithms
    • cross validation


    Dive into the research topics of 'No free lunch for cross validation'. Together they form a unique fingerprint.

    Cite this