Presentation | 2013-01-24 Analysis of Medical Treatment Data using Inverse Reinforcement Learning Hideki ASOH, Masanori SHIRO, Toshihiro KAMISHIMA, Shotaro AKAHO, Takahide KOHRO, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | It is an important issue to utilize large amount of medical records which are accumulated on medical information systems. Medical treatment processes can be considered as interaction processes between doctors and a patients. From the viewpoint, we are modeling medical treatment records by Markov decision processes. By the method, we can simulate future states of patients and evaluate the value of medical treatments. In order to evaluate the value of treatment, reward function which doctors are using is necessary. However, treatment records do not include the reward information. To cope with such problem, the problem of inverse reinforcement learning (IRL) has been formalized and various algorithms to solve the problem are proposed. IRL estimates experts' reward function from experts' observation and action data. In this presentation, we introduce IRL and report on the applicability of the Bayesian IRL to the medical records of heart disease patients. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Inverse Reinforcement Learning / Markov Decision Process / Data Mining / Medical Record |
Paper # | NLP2012-106,NC2012-96 |
Date of Issue |
Conference Information | |
Committee | NLP |
---|---|
Conference Date | 2013/1/17(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 | Nonlinear Problems (NLP) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Analysis of Medical Treatment Data using Inverse Reinforcement Learning |
Sub Title (in English) | |
Keyword(1) | Inverse Reinforcement Learning |
Keyword(2) | Markov Decision Process |
Keyword(3) | Data Mining |
Keyword(4) | Medical Record |
1st Author's Name | Hideki ASOH |
1st Author's Affiliation | National Institute of Advanced Industrial Science and Technology (AIST)() |
2nd Author's Name | Masanori SHIRO |
2nd Author's Affiliation | National Institute of Advanced Industrial Science and Technology (AIST) |
3rd Author's Name | Toshihiro KAMISHIMA |
3rd Author's Affiliation | National Institute of Advanced Industrial Science and Technology (AIST) |
4th Author's Name | Shotaro AKAHO |
4th Author's Affiliation | National Institute of Advanced Industrial Science and Technology (AIST) |
5th Author's Name | Takahide KOHRO |
5th Author's Affiliation | The University of Tokyo Hospital |
Date | 2013-01-24 |
Paper # | NLP2012-106,NC2012-96 |
Volume (vol) | vol.112 |
Number (no) | 389 |
Page | pp.pp.- |
#Pages | 5 |
Date of Issue |