Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
You can not select more than 25 topics
Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
LWXuan
86fc0a56fe
|
3 years ago | |
|---|---|---|
| .. | ||
| pegasos | 3 years ago | |
| pegasosMultLambda | 3 years ago | |
| resultFiles | 3 years ago | |
| README | 3 years ago | |
| binarymkl.m | 3 years ago | |
| construct_targetmat.m | 3 years ago | |
| mybinarymkl.m | 3 years ago | |
| paths.m | 3 years ago | |
| read_model.m | 3 years ago | |
| write_svm.m | 3 years ago | |
| write_svm_trval.m | 3 years ago | |
README
Paper: A binary classification framework for two stage multiple kernel learning, ICML 2011
Usage:
[w1 w2] = binarymkl (M, tr_label, pathfile, resultfile, EXPname, permind);
(read the comments in file binarymkl.m for description of input and output parameters)
Contents:
- Two pegasos based implementations (modified from original pegasos to add additional projection onto positive orthant)
-- pegasos: do "make" to build the executable on your platform
-- pegasosMultLambda: do "make" to build the executable on your platform
- paths.m: example pathfile
- other .m files needed for running "binarymmkl.m"
References:
pegasos:
Shai Shalev-Shwartz, Yoram Singer, Nathan Srebro, "Pegasos: Primal Estimated sub-GrAdient SOlver for SVM", ICML 2007
An efficient radius-incorporated MKL algorithm for Alzheimer’s disease prediction
C++ MATLAB Python Jupyter Notebook SWIG other