Presentation | 2011-03-29 カーネル密度比推定の統計的解析(学習問題の解析,テキスト・Webマイニング,一般) Takafumi Kanamori, Taiji Suzuki, Masashi Sugiyama, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | The ratio of two probability densities can be used for solving various machine learning tasks such as covariate shift adaptation (importance sampling), outlier detection (likelihood-ratio test), feature selection (mutual information), and conditional probability estimation. Several methods of directly estimating the density ratio have been recently developed, e.g., moment matching estimation, maximum-likelihood density-ratio estimation, and least-squares density-ratio fitting. In this paper, we propose a kernelized variant of the least-squares method for density-ratio estimation, which is called kernel unconstraint least-squares importance fitting (KuLSIF). We then investigate its fundamental statistical properties including a non-parametric convergence rate, an analytic-form solution and a leave-one-out cross-validation score. We further study its relation to other kernel-based density-ratio estimators. In experiments, we numerically compare various kernel-based density-ratio estimation methods, and show that KuLSIF compares favorably with other approaches. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | density ratio / kernel method / statistical consistency |
Paper # | IBISML2010-110 |
Date of Issue |
Conference Information | |
Committee | IBISML |
---|---|
Conference Date | 2011/3/21(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Information-Based Induction Sciences and Machine Learning (IBISML) |
---|---|
Language | ENG |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | |
Sub Title (in English) | |
Keyword(1) | density ratio |
Keyword(2) | kernel method |
Keyword(3) | statistical consistency |
1st Author's Name | Takafumi Kanamori |
1st Author's Affiliation | Nagoya University() |
2nd Author's Name | Taiji Suzuki |
2nd Author's Affiliation | University of Tokyo |
3rd Author's Name | Masashi Sugiyama |
3rd Author's Affiliation | Tokyo Institute of Technology |
Date | 2011-03-29 |
Paper # | IBISML2010-110 |
Volume (vol) | vol.110 |
Number (no) | 476 |
Page | pp.pp.- |
#Pages | 8 |
Date of Issue |