Presentation | 2002/1/22 Mixture of Asymmetric Gaussians and Its Maximum Likelihood Estimation Tsuyoshi KATO, Shinichiro OMACHI, Hirotomo ASO, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In this paper, we propose a new probability model, 'asymmetric Gaussian(AG)', which is an extension of Gaussian. The AG can capture spatially asymmetric distributions, yet remaining computationally tractable. There is a similar work [1], which introduces 'Asymmetric Mahalanobis Distance (AMD)' and applys it to handwritten Chinese and Japanese character recognition. The AMD can measure a spatially asymmetrical distance between an unknown pattern and a category class and shows excellent classification performance. However, the AMD is suitable for only an unimodal distribution, so the range of its application is necessarily somewhat limited. Meanwhile, since our model is formulated by a density function, it is easy to be extended to mixture of asymmetric Gaussians (MAG), which can capture multimodal distributions. We apply the AGs to character classification problem and show that the AGs outperform Gaussian models. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Asymmetric Gaussian / Mixture of Asymmetric Gaussians / Expectation-Conditional Maximization algorithm / Character recognition |
Paper # | |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2002/1/22(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 | Neurocomputing (NC) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Mixture of Asymmetric Gaussians and Its Maximum Likelihood Estimation |
Sub Title (in English) | |
Keyword(1) | Asymmetric Gaussian |
Keyword(2) | Mixture of Asymmetric Gaussians |
Keyword(3) | Expectation-Conditional Maximization algorithm |
Keyword(4) | Character recognition |
1st Author's Name | Tsuyoshi KATO |
1st Author's Affiliation | Graduate School of Engineering, Tohoku University() |
2nd Author's Name | Shinichiro OMACHI |
2nd Author's Affiliation | Graduate School of Engineering, Tohoku University |
3rd Author's Name | Hirotomo ASO |
3rd Author's Affiliation | Graduate School of Engineering, Tohoku University |
Date | 2002/1/22 |
Paper # | |
Volume (vol) | vol.101 |
Number (no) | 616 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |