This task is about translating paper summaries into doxygen comments.
We have a lot of SVM implementations in Shogun, see http://www.shogun-toolbox.org/doc/en/latest/classshogun_1_1CSVM.html
This diversity is great since it offers a specific solver for a large range of problem classes. The problem here is to know which SVM to use in which context. The current lack of @class descriptions does not help.
The task is to make sure that every SVM implementation in Shogun has
A proper list of reference papers, in a proper citation format (i.e. not just the link to the pdf, but something that would appear in a paper reference list)
A brief description what the solver is doing, maybe based on the abstract of the main reference
A description of optimal use cases. Where should/can the SVM be used? What are its strengths/weaknesses. What are caveats when using it? What are alternatives?
The above also should contain a rough indication of the computational costs.
Be sure to include most important mathematical objects in latex, such as objective functions, etc
If you get stuck with any of those points, or cannot find the references, ask @sonney2k He is the SVM expert, knows all about the above questions, and therefore assigned to this issue :)
This task is about translating paper summaries into doxygen comments.
We have a lot of SVM implementations in Shogun, see http://www.shogun-toolbox.org/doc/en/latest/classshogun_1_1CSVM.html This diversity is great since it offers a specific solver for a large range of problem classes. The problem here is to know which SVM to use in which context. The current lack of @class descriptions does not help.
The task is to make sure that every SVM implementation in Shogun has
If you get stuck with any of those points, or cannot find the references, ask @sonney2k He is the SVM expert, knows all about the above questions, and therefore assigned to this issue :)