Department of Chemical and Biomolecular Engineering
Korea Advanced Institute of Science and Technology

제목

12월 첫수융합포럼 안내 The First Wednesday Multidisciplinary Forum, December, 2012

 

12월 첫수융합포럼 안내 The First Wednesday Multidisciplinary Forum, December, 2012

 

첫수 융합포럼 개최 안내

The First Wednesday Multidisciplinary Forum

 

문화기술대학원, 산업 시스템 공학과, 지식서비스공학과 3 학과 공동으로 주관하는

12 첫수 융합포럼 여러분을 초대합니다.

Graduate School of Culture Technology, Department of Industrial & Systems Engineering,

and Department of Knowledge Service Engineering would like to invite KAIST students to

"The First Wednesday Multidisciplinary Forum on December 5(Wed), 2012.

 

 

일시 : 2012. 12. 5() 정오12:00

Date : December 5(Wed), 2012, 12:00 at noon

 

장소 : 창의학습관(Bldg# E11) 101

Venue : Room# 101, Creative Learning Bldg. (#E11)

 

연사 발표 주제

학과명 Dept.

연사 Speaker

발표 주제 Topic to present

문화기술대학원

Graduate School of Culture Technology, KAIST

박주용 교수

Prof. Park, juyong

Inference of rankings from data: Bayesian-approach to determining dominance hierarchy

산업 시스템 공학과

Department of Industrial & Systems Engineering, KAIST

문일철 교수

Prof. Moon, Il Chul

Predicting the Stock Market with Unstructured Media Texts

지식서비스공학과

Department of Knowledge Service Engineering, KAIST

권가진 교수

Prof. Gweon, Gahgene

Cracking the code in social interactions

 

Abstract

Inference of rankings from data: Bayesian-approach to determining dominance

hierarchy

Bayes' theorem, first formulated by the British mathematician Thomas Bayes in the 18th

century, still functions as the standard methodology by which one's distribution of a

parameter is updated according to the unearthing of new empirical evidence. Here we

apply this method to the problem of ranking nodes in a competition network, which

produces a confidence interval in the rankings, unlike centrality-based methods.

 

Predicting the Stock Market with Unstructured Media Texts

I present the result of predicting the stock market index with unstructured media texts.

The stock market index is a single measurement of a collective behavior of investors,

and from such perspective, I conjecture that the collective would be better explained

by utilizing the media data generated by the individuals in the collective. I compare

and contrast the prediction accuracy and the values of the traditional economic

indicators as well as the recent unstructured text-mining data.

 

Cracking the code in social interactions

My general research goal is in automatically analyzing various social interactions in

order to gain understanding on social phenomena and using that understanding to

improve quality of life (e.g. education, relationships). Analyzing and making sense of

social interactions, more specifically conversations, is analogous to cracking the code

thats hidden under seemingly ordinary conversations. In this talk, I highlight the

necessity of such research along with some example research work in this area.

            

간단한 점심식사를 준비할 예정입니다.

Refreshments will be served for free.

 

관련문의 /Inquiry :

문화기술대학원 이윤정 / Ext. 2902, yclee@kaist.ac.kr (Mrs. Lee, Y. J.)

산업 시스템 공학과 노수정 / Ext. 3102, nohsj@kaist.ac.kr (Mrs. Noh, S. J.)

지식서비스공학과 최진희 / Ext. 1603, choijh84@kaist.ac.kr (Mrs. Lee, H. M.)

첨부파일 다운로드

등록자관리자

등록일2012-12-04

조회수6,461

  • 페이스북 공유
  • 트위터 공유
  • Google+ 공유
  • 인쇄하기