| 학회 | 한국화학공학회 |
| 학술대회 | 2014년 가을 (10/22 ~ 10/24, 대전 DCC) |
| 권호 | 20권 2호, p.1298 |
| 발표분야 | 공정시스템 |
| 제목 | feature selection in supervised learning in an spectroscopic problem |
| 초록 | The need to identify a few important variables that affect a certain outcome of interest commonly arises in all research areas. Moreover it is convenient to reduce the number of involved features in order to reduce the complexity, using feature selection methods as powerful tools in analysis of high dimensional massive data, which plays more important role when the number of variables (p) far exceeds the number of observations (n), which makes the traditional statistical methods infeasible for data analysis. This paper presents comparison of variables selection methods such as LARS, Lasso, GA and also traditional methods such as forward and stepwise selection, in a soil carbonate determination case study on FT-IR and XDR data as high dimensional datasets, to extract important variables and predict soil carbonate using SVM. The results show high ability of LARS and Lasso in extracting effective variables on carbonate determination and the least prediction error was obtained by applying subset by LARS. |
| 저자 | FarnazPirasteh, 유 준 |
| 소속 | 부경대 |
| 키워드 | feature selection; data mining; spectroscopy |
| VOD | VOD 보기 |
| 원문파일 | 초록 보기 |
