Comparison of Boosting and SVM
Comparison of Boosting and SVM
- 주제(키워드) boosting , classification , prediction , SVM , boosting , classification , prediction , SVM
- 발행기관 한국데이터정보과학회
- 발행년도 2005
- 총서유형 Journal
- UCI G704-000605.2005.16.4.019
- KCI ID ART001131297
초록/요약
We compare two popular algorithms in current machine learning and statistical learning areas, boosting method represented by AdaBoost and kernel based SVM (Support Vector Machine) using 13 real data sets. This comparative study shows that boosting method has smaller prediction error in data with heavy noise, whereas SVM has smaller prediction error in the data with little noise.
more