Kryterium Akaike: prostota w języku statystyki
Akaike Criterion: Simplicity in the Language of Statistics
Author(s): Łukasz Kukier, Marek Szydłowski, Paweł TamborSubject(s): Philosophy
Published by: Towarzystwo Naukowe KUL & Katolicki Uniwersytet Lubelski Jana Pawła II
Keywords: Akaike Informative Criterion; simplicity; philosophy of science; model selection; kryterium Akaike; prostota; filozofia nauki; kryteria selekcji modeli
Summary/Abstract: Many authors have pointed out that notion of simplicity is unclear. For deeper understanding of this term, we investigate it in the conceptual framework of Akaike Informative Criterion (AIC). Advantages as well as troubles of such formulation are presented in the paper. We also discuss a role which plays the simplicity notion formulated in statistical framework within Sober’s philosophy of science conception.
Journal: Roczniki Filozoficzne
- Issue Year: 57/2009
- Issue No: 1
- Page Range: 91-126
- Page Count: 36
- Language: Polish