The document summarizes support vector machines (SVM), a supervised machine learning method for classification and regression. It discusses that SVM finds the optimal separating hyperplane between two classes with maximum margin. It also covers modifications for non-linearly separable data and multiclass classification problems using one-vs-rest and one-vs-one approaches. An example applying SVM to classify iris data is provided. In summary, the document provides an overview of SVM including its goal of finding optimal decision boundaries, extensions for complex problems, and an illustrative example.
1. Support Vector Machine
Putri W Novianti
Victor L Jong
Biostatistics & Research Support
Julius Center for Health Sciences and Primary Care
University Medical Center Utrecht
2. Support Vector Machine 2
• Binary classification method
• The method finds the best decision hyperplane that separate sample from
two classes with maximum margin
19. Support Vector Machine 19
Multiclass outcome
- SVM only handle binary classification
- Although binary classification is the most common classification in
microarray, multiclass outcome could be occur in practice
- Modification is needed to handle multiclass outcome
- one-versus-rest (OVR)
- one-versus-one (OVO)
[2]
32. Support Vector Machine 32
References
[1] Zhang, X. Support Vector Machine. Lecture slides on Data Mining course. Fall 2010, KSA: KAUST
[2] Statnikov, A. et al. 2005. A comprehensive evaluation of multicategory classification methods for microarray gene expression cancer
diagnosis. Bioinformatics, 21:5, 631-643
[3] Hastie, T., Tibshirani, R., Friedman, J. The elements of statistical learning, second edition. 2009. New York: Springer
[4] Guyon, I et al. 2002. Gene selection for cancer classification using support vector machines. Machine Learning, 49, 389-422
[5] Meyer, D. et al. 2012. R package: e1071.