Wallenius Naive Bayes

Enric Junque de Fortuny; D Martens; F Provost

Wallenius Naive Bayes

Enric Junque de Fortuny, D Martens, F Provost

Department of Marketing Management

Research output: Chapter/Conference proceeding › Conference proceeding › Academic › peer-review

Abstract

Traditional event models underlying naive Bayes classifiers assume probability distributions that are not appropriate for binary data generated by human behaviour. In this work, we develop a new event model, based on a somewhat forgotten distribution created by Kenneth Ted Wallenius in 1963. We show that it achieves superior performance using less data on a collection of Facebook datasets, where the task is to predict personality traits, based on likes.

Original language	English
Title of host publication	-
Publication status	In preparation - 2019

Research programs

RSM MKT

Access to Document

https://ssrn.com/abstract=2370823

Cite this

@inproceedings{93771801230840eeb2dcd3db7b7a9a00,

title = "Wallenius Naive Bayes",

abstract = "Traditional event models underlying naive Bayes classifiers assume probability distributions that are not appropriate for binary data generated by human behaviour. In this work, we develop a new event model, based on a somewhat forgotten distribution created by Kenneth Ted Wallenius in 1963. We show that it achieves superior performance using less data on a collection of Facebook datasets, where the task is to predict personality traits, based on likes.",

author = "{Junque de Fortuny}, Enric and D Martens and F Provost",

year = "2019",

language = "English",

booktitle = "-",

}

TY - GEN

T1 - Wallenius Naive Bayes

AU - Junque de Fortuny, Enric

AU - Martens, D

AU - Provost, F

PY - 2019

Y1 - 2019

N2 - Traditional event models underlying naive Bayes classifiers assume probability distributions that are not appropriate for binary data generated by human behaviour. In this work, we develop a new event model, based on a somewhat forgotten distribution created by Kenneth Ted Wallenius in 1963. We show that it achieves superior performance using less data on a collection of Facebook datasets, where the task is to predict personality traits, based on likes.

AB - Traditional event models underlying naive Bayes classifiers assume probability distributions that are not appropriate for binary data generated by human behaviour. In this work, we develop a new event model, based on a somewhat forgotten distribution created by Kenneth Ted Wallenius in 1963. We show that it achieves superior performance using less data on a collection of Facebook datasets, where the task is to predict personality traits, based on likes.

M3 - Conference proceeding

BT - -

ER -

Wallenius Naive Bayes

Abstract

Research programs

Access to Document

Fingerprint

Cite this