Linear Probability Forecasting - Artificial Intelligence Applications and Innovations Access content directly
Conference Papers Year : 2010

Linear Probability Forecasting

Abstract

In this paper we consider two online multi-class classification problems: classification with linear models and with kernelized models. The predictions can be thought of as probability distributions. The quality of predictions is measured by the Brier loss function. We suggest two computationally efficient algorithms to work with these problems, the second algorithm is derived by considering a new class of linear prediction models. We prove theoretical guarantees on the cumulative losses of the algorithms. We kernelize one of the algorithms and prove theoretical guarantees on the loss of the kernelized version. We perform experiments and compare our algorithms with logistic regression.
Fichier principal
Vignette du fichier
ZhdanovK10.pdf (177.25 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01060645 , version 1 (16-11-2017)

Licence

Attribution

Identifiers

Cite

Fedor Zhdanov, Yuri Kalnishkan. Linear Probability Forecasting. 6th IFIP WG 12.5 International Conference on Artificial Intelligence Applications and Innovations (AIAI), Oct 2010, Larnaca, Cyprus. pp.4-11, ⟨10.1007/978-3-642-16239-8_4⟩. ⟨hal-01060645⟩
130 View
70 Download

Altmetric

Share

Gmail Facebook X LinkedIn More