Direct Density-Ratio Estimation with Dimensionality Reduction via Hetero-Distributional Subspace Analysis
Abstract
Methods for estimating the ratio of two probability density functions have been actively explored recently since they can be used for various data processing tasks such as non-stationarity adaptation, outlier detection, feature selection, and conditional probability estimation. In this paper, we propose a new density-ratio estimator which incorporates dimensionality reduction into the density-ratio estimation procedure. Through experiments, the proposed method is shown to compare favorably with existing density-ratio estimators in terms of both accuracy and computational costs.
Cite
Text
Yamada and Sugiyama. "Direct Density-Ratio Estimation with Dimensionality Reduction via Hetero-Distributional Subspace Analysis." AAAI Conference on Artificial Intelligence, 2011. doi:10.1609/AAAI.V25I1.7905Markdown
[Yamada and Sugiyama. "Direct Density-Ratio Estimation with Dimensionality Reduction via Hetero-Distributional Subspace Analysis." AAAI Conference on Artificial Intelligence, 2011.](https://mlanthology.org/aaai/2011/yamada2011aaai-direct/) doi:10.1609/AAAI.V25I1.7905BibTeX
@inproceedings{yamada2011aaai-direct,
title = {{Direct Density-Ratio Estimation with Dimensionality Reduction via Hetero-Distributional Subspace Analysis}},
author = {Yamada, Makoto and Sugiyama, Masashi},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2011},
pages = {549-554},
doi = {10.1609/AAAI.V25I1.7905},
url = {https://mlanthology.org/aaai/2011/yamada2011aaai-direct/}
}