The Observation Theory EncyclopediaFrom TSKAboutBy kindBy chapterBy Lean fileLedgerProvenance

ensemble instrument

DefinitionA scorer that averages or votes several scorers. Its average lies between its members and its squared error is at most their mean squared error. Chapter 7. Also bagging, random forest.
ExampleScores 0.6, 0.7, and 0.8 average 0.7, and the average’s squared error is at most the mean of the three squared errors.
BookData Mining as Observation, draft 0.2, commit f3914f0; entry id ensemble, kind instrument.
Statusno ledger row names this entry. Corrections: none recorded.
Defining equationnone
Assumptions and scope
  • A scorer that averages or votes several scorers. Its average lies between the smallest and the largest member, its squared error is at most the mean of the members’ squared errors, and a constant ensemble is its constant. That arithmetic is why bagging reduces variance.
  • Whether an ensemble beats a formula is a measured question and not a principle, and the comparison is fair only inside one fold protocol with the Nadeau and Bengio correction applied, where nine wins to three became three wins, seventeen ties, and eleven losses.
Prior artnone recorded
Evidencelean/DataMiningAsObservation/Ensemble.lean
Reviewednot yet reviewed; generated 2026-09-10 from records at the commits on the provenance page.
00.51estimateresamples of the same rows
An average of several scorers, between its members.

Equation

none

Conditions

Conditions are curated in entries.toml rather than read from a record.

Ledger

none

First stated

Breiman, bagging predictors, 1996, as chapter 7 section 7.1 of Data Mining as Observation reads it, with the fair comparison in constraint-gap/review/INDETERMINATES.md:1-40.

Measurements

Where the book states it Numbers, as the book’s sources table records them Source
chapter 7 section 7.1 bagging variance, out-of-bag estimation, random forests, importances, AdaBoost Hastie, Tibshirani, Friedman, ESL 2e chapters 15 and 10.1; TSK 2e 4.10

Failures and corrections

none

Invariance envelope

none declared

Machine checked

lean/DataMiningAsObservation/Ensemble.lean, theorems average_between, sq_average_le, average_const, at observation-data-mining f3914f0; what the check covers is stated in the book’s appendix C.

Used in

Data Mining as Observation chapters 0, 6, 7.

Related

boosting; decision tree; standard error; Nadeau and Bengio correction.

See also

Book equations stated beside the entry’s terms, not defining it: 6.1, 0.18, 8.2.

Sources-table rows that share a record with the entry without naming it: chapter 6 section 6.3, chapter 7 section 7.3, chapter 7 section 7.4.

Status

Generated 2026-09-10 by encyclopedia/generate.py; book at observation-data-mining f3914f0; the commit of every record is listed in the encyclopedia’s provenance.

← encoderentropy →