Avoiding biases in binned fits. (4th August 2021)

• Record Type: Journal Article
• Authors: Gligorov, V.V.; Hageboeck, S.; Nanut, T.; Sciandra, A.; Tou, D.Y.
• Is Part Of: Journal of instrumentation. Volume 16:Number 8(2021)
• Publication Date: 2021-08-04
• Subjects: Analysis and statistical methods -- Data processing methods; Scientific apparatus and instruments -- Periodicals; 502.84
• DOI: 10.1088/1748-0221/16/08/T08004
• Languages: English
• ISSNs: 1748-0221

Abstract

Abstract: Binned maximum likelihood fits are an attractive option when analysing large datasets, but require care when computing likelihoods of continuous PDFs in bins. For many years the widely used statistical modelling package evaluated probabilities at the bin centre, leading to significant biases for strongly curved probability density functions. We demonstrate the biases with real-world examples, and introduce a PDF class to that removes these biases. The physics and computation performance of this new class are discussed.