DESCRIPTION

---

Class for a Nu-SVM classifier. This SVM type uses a Nu parameter in order to manage the error function. The supported algorithm in BioClass are based on LibSVM library[1][2]

PARAMETERS

---

Classifier Name: Element identifier for operations and Clipboard.

Nu: Nu parameter value. It permits to minimize the error function. Only float values between 0.0 to 0.5 are allowed.

Tolerance of Termination: Tolerance value for the termination criterion algorithm.

Kernel Type: Kernel type to use, is possible to select between: Linear, Poplynomial, Radial or Sigmoid.

Normalize: Whether to normalize the data.

Probability Estimates: Whether to generate probability estimates instead of [-1/+1] prediction values. It reduces the classification errors in some cases.

Shrinking: if this parameter is enable, SVM uses the shrinking heuristic in order to improve the search algorithm.

REFERENCES

---

[1]Yasser EL-Manzalawy (2005). WLSVM. URL http://www.cs.iastate.edu/~yasser/wlsvm/.

[2]Chih-Chung Chang, Chih-Jen Lin (2001). LIBSVM - A Library for Support Vector Machines. URL http://www.csie.ntu.edu.tw/~cjlin/libsvm/