Paper Abstract and Keywords |
Presentation |
2018-11-05 15:10
[Poster Presentation]
In-Hospital Mortality Prediction for Trauma Patients Using Cost-Sensitive MedLDA Haruya Ishizuka, Tsukasa Ishigaki, Naoya Kobayashi, Daisuke Kudo, Atsuhiro Nakagawa (Tohoku Univ.) IBISML2018-75 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
In the field of emergency medical care, decision support systems, which support doctors’ judgements, have attracted. Accurate mortality prediction is essential to realize such systems. Recently, some researchers have exploited outcome prediction methods using clinical texts. Supervised topic models (sTMs) are known as discriminative models using textual information, reveal high predictive performance in some tasks. However, previous sTMs give the same amount of misclassification penalty both of majority and minority instances. In-hospital mortality rate is about 10$%$, one example of imbalanced data. The ability of sTMs to detect severe ill patients may degrade due to imbalanced data problem. To address this issue, we propose Cost-Sensitive MedLDA in this paper. Cost-Sensitive MedLDA is an sTM which estimate the latent topic representation and discriminant function jointly using max-margin principle and cost-sensitive learning. Through mortality prediction experiments, we show that the proposed model can predict patients’ outcome more accurately compared with previous methods. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
supervised topic model / imbalanced data / cost-sensitive learning / mortality prediction / medical data analysis / / / |
Reference Info. |
IEICE Tech. Rep., vol. 118, no. 284, IBISML2018-75, pp. 233-240, Nov. 2018. |
Paper # |
IBISML2018-75 |
Date of Issue |
2018-10-29 (IBISML) |
ISSN |
Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
IBISML2018-75 |
Conference Information |
Committee |
IBISML |
Conference Date |
2018-11-05 - 2018-11-07 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
Hokkaido Citizens Activites Center (Kaderu 2.7) |
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
Information-Based Induction Science Workshop (IBIS2018) |
Paper Information |
Registration To |
IBISML |
Conference Code |
2018-11-IBISML |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
In-Hospital Mortality Prediction for Trauma Patients Using Cost-Sensitive MedLDA |
Sub Title (in English) |
|
Keyword(1) |
supervised topic model |
Keyword(2) |
imbalanced data |
Keyword(3) |
cost-sensitive learning |
Keyword(4) |
mortality prediction |
Keyword(5) |
medical data analysis |
Keyword(6) |
|
Keyword(7) |
|
Keyword(8) |
|
1st Author's Name |
Haruya Ishizuka |
1st Author's Affiliation |
Tohoku University (Tohoku Univ.) |
2nd Author's Name |
Tsukasa Ishigaki |
2nd Author's Affiliation |
Tohoku University (Tohoku Univ.) |
3rd Author's Name |
Naoya Kobayashi |
3rd Author's Affiliation |
Tohoku University (Tohoku Univ.) |
4th Author's Name |
Daisuke Kudo |
4th Author's Affiliation |
Tohoku University (Tohoku Univ.) |
5th Author's Name |
Atsuhiro Nakagawa |
5th Author's Affiliation |
Tohoku University (Tohoku Univ.) |
6th Author's Name |
|
6th Author's Affiliation |
() |
7th Author's Name |
|
7th Author's Affiliation |
() |
8th Author's Name |
|
8th Author's Affiliation |
() |
9th Author's Name |
|
9th Author's Affiliation |
() |
10th Author's Name |
|
10th Author's Affiliation |
() |
11th Author's Name |
|
11th Author's Affiliation |
() |
12th Author's Name |
|
12th Author's Affiliation |
() |
13th Author's Name |
|
13th Author's Affiliation |
() |
14th Author's Name |
|
14th Author's Affiliation |
() |
15th Author's Name |
|
15th Author's Affiliation |
() |
16th Author's Name |
|
16th Author's Affiliation |
() |
17th Author's Name |
|
17th Author's Affiliation |
() |
18th Author's Name |
|
18th Author's Affiliation |
() |
19th Author's Name |
|
19th Author's Affiliation |
() |
20th Author's Name |
|
20th Author's Affiliation |
() |
Speaker |
Author-1 |
Date Time |
2018-11-05 15:10:00 |
Presentation Time |
180 minutes |
Registration for |
IBISML |
Paper # |
IBISML2018-75 |
Volume (vol) |
vol.118 |
Number (no) |
no.284 |
Page |
pp.233-240 |
#Pages |
8 |
Date of Issue |
2018-10-29 (IBISML) |
|