isip_svm_learn -p param.sof [options]

isip_svm_learn -help;
isip_svm_learn -params_file params.sof -pos_examples pos_examples.sof
               -neg_examples neg_examples.sof

This utility computes the support vector model from a set of in-class and out-of-class observations.

• -help: displays a help message
• -verbose: log verbosity level
  • NONE: output the number of files processed
  • BRIEF: output also names of the files processed
  • DETAILED: output also hypotheses, scores and frames number
  
  
• -debug_level: debugging information level (default NONE)
  • NONE: output the number of files processed
  • BRIEF: output also names of the files processed
  • DETAILED: output also hypotheses, scores and frames number
  
  
• -params_file: support vector model parameter file
• -pos_examples: in-class observations identifiers
• -neg_examples: out-of-class observations identifiers
• -penalty: training error vs margin trade-off parameter (default 50)
• -epsilon: allowable error for QP termination criterion (default 0.001)
• -tolerance: inaccuracies allowed around the KKT conditions (default 0.001)
• -kappa: kappa parameter of the sigmoid kernel (default 1.0)
• -delta: delta parameter of the sigmoid kernel (default 1.0)
• -degree: degree parameter of the polynomial kernel (default 3.0)
• -gamma: gamma parameter of the rbf kernel (default 0.5)
  
  
• -log_file: log file used instead of stdout

• None.