MATLAB (and custom-built toolboxes/library usually freely available) has been used to implement FSVM, such as in "Using Fuzzy Support Vector Machine Network to Predict Low Homology Protein Structural Classes" (PRIB 2007), "Fuzzy Chance Constrained Support Vector Machine" (LSMS 2010), "Detection for Abnormal Event Based on Trajectory Analysis and FSVM" (ICIC 2007), etc. The most common toolboxes I've seen are LibSVM (http://www.csie.ntu.edu.tw/~cjlin/libsvm/), LS-SVMlab (http://www.esat.kuleuven.be/sista/lssvmlab/toolbox.html) FSTB or (more recently) the fuzzy logic toolbox (http://www.mathworks.com/products/fuzzy-logic/), but mostly custom code specific to the paper/project in question and created by the authors. LibSVM has a textbook that mentions fuzzy methods in passing (p. 235), and FSTB has guidebooks that are out there but outdated (http://www.lasc.univ-metz.fr/IMG/pdf/fstb-reference.pdf). MATLAB has the advantage over SymbolicC++, C#, etc., in that even without some set of .m files for fuzzy system/set functions, MATLAB is designed around matrices and vectors while SymbolicC++, Java, C#, etc., are neither designed for matrix manipulations nor for mathematical functions in general (SymbolicC++ is, but only as an add-on that turns C++ into a something that is still more difficult to use than MATLAB for modelling, analysis, statistical tests, etc.).
MATLAB (and custom-built toolboxes/library usually freely available) has been used to implement FSVM, such as in "Using Fuzzy Support Vector Machine Network to Predict Low Homology Protein Structural Classes" (PRIB 2007), "Fuzzy Chance Constrained Support Vector Machine" (LSMS 2010), "Detection for Abnormal Event Based on Trajectory Analysis and FSVM" (ICIC 2007), etc. The most common toolboxes I've seen are LibSVM (http://www.csie.ntu.edu.tw/~cjlin/libsvm/), LS-SVMlab (http://www.esat.kuleuven.be/sista/lssvmlab/toolbox.html) FSTB or (more recently) the fuzzy logic toolbox (http://www.mathworks.com/products/fuzzy-logic/), but mostly custom code specific to the paper/project in question and created by the authors. LibSVM has a textbook that mentions fuzzy methods in passing (p. 235), and FSTB has guidebooks that are out there but outdated (http://www.lasc.univ-metz.fr/IMG/pdf/fstb-reference.pdf). MATLAB has the advantage over SymbolicC++, C#, etc., in that even without some set of .m files for fuzzy system/set functions, MATLAB is designed around matrices and vectors while SymbolicC++, Java, C#, etc., are neither designed for matrix manipulations nor for mathematical functions in general (SymbolicC++ is, but only as an add-on that turns C++ into a something that is still more difficult to use than MATLAB for modelling, analysis, statistical tests, etc.).
Read this book
LEARNING AND SOFT COMPUTING
Support Vector Machines, Neural Networks and Fuzzy Logic Models
http://www.support-vector.ws/html/downloads.html
Any idea of IT2FLS + SVM? I need to compare one of my model with IT2 FLS + any model.
- Badges
- Science topic