國立政治大學應用數學系學 術 演 講
時 間：97年5月5日(星期一) 14:10 – 15:00
演講人：Arjun K. Gupta (Bowling Green State University)
題 目：Information Theory : Detection of Change and Data Mining
摘 要：Both Statistics and Data mining are concerned with drtawing inferences from data.The aim of inference may be understanding the pattern of correlation and causal links among the data values ( explanation ) or making predictions for the data values ( generalizations ).Whereas data mining avoids the use of models ,statistics has developed an approach that involves specifying a model for the probability distribution of the data and making inferences in the form of probability statements.In this talk, we propose a method for detecting changes (pattern) in the data .The method transforms the problem to a model selection problem and uses Akaike's Information Criterion ( AIC or SIC ).The number of changes are detected and the change points are estimated consistently .The methodology is successfully illustrated by applying to the New York Stock Exchange stock market data.
地 點：志希二樓 070221
備 註：茶會時間為13:30 ~ 14:00於教授休息室。