Retired

This document has been retired.

---

Fast Logistic Regression for Data Mining, Text Classification and Link Detection (2003)

Tags

Life Science Data Mining, Logistic Regression

Abstract

Previous work by the authors (Komarek, AISTAT 2003) demonstrated that logistic regression can be a fast and accurate data mining tool for life sciences datasets, competitive with modern tools like support vector machines and ball-tree based K-NN. This paper has two objectives. The first objective is a serious empirical comparison of logistic regression to several classical and modern learners on a variety of learning tasks. The second is to describe our use of conjugate gradient inside an iteratively re-weighted least squares fitting procedure.

Full text

Download (application/pdf, 78.3 kB)

Approximate BibTeX Entry

@article{komarek:nips2003,
    Howpublished = {Submitted to NIPS 2003},
    Month = {June},
    Year = {2003},
    Booktitle = {Submitted to NIPS},
    Title = {Fast Logistic Regression for Data Mining, Text Classification and Link Detection}
}