OnevsOne updated library?

[stallam-unb]

Hello,

I am trying to implement OvO multiclass logistic regression classifier in MATLAB, but the version on the webpage for the OvO code is older than the current version of LIBLINEAR (2.20 vs 2.30). Would it be possible to obtain an updated version for this code? Thanks in advance.

[cjlin1]

The latest version will be uploaded in one month. Thanks

On 2019-07-20 07:56, stpr-unb wrote:

Hello,

I am trying to implement OvO multiclass logistic regression classifier in MATLAB, but the version on the webpage [1] for the OvO code is older than the current version of LIBLINEAR (2.20 vs 2.30). Would it be possible to obtain an updated version for this code? Thanks in advance.

-- You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub [2], or mute the thread [3]. [ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "https://github.com/cjlin1/liblinear/issues/54?email_source=notifications\u0026email_token=ABI3BHVSHFAR7O2D7MDA7LTQAMRTVA5CNFSM4IFO3ZN2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HAOBXIA", "url": "https://github.com/cjlin1/liblinear/issues/54?email_source=notifications\u0026email_token=ABI3BHVSHFAR7O2D7MDA7LTQAMRTVA5CNFSM4IFO3ZN2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HAOBXIA", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } } ]

Links:

[1] https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/#liblinear_for_one_versus_one_multi_class_classification [2] https://github.com/cjlin1/liblinear/issues/54?email_source=notifications&email_token=ABI3BHVSHFAR7O2D7MDA7LTQAMRTVA5CNFSM4IFO3ZN2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HAOBXIA [3] https://github.com/notifications/unsubscribe-auth/ABI3BHWVYC4HSMO7P55JYLDQAMRTVANCNFSM4IFO3ZNQ

[stallam-unb]

Thank you. A follow up question - I should be able to drop it into the LIBLINEAR directory and compile the mex file as I would for the standard LIBLINEAR correct?

[cjlin1]

In the released ovo extension the matlab directory was removed because we didn't intend to maintain it.. But I think we have the code (note that things in the master cannot be used due to the change on the number of w vectors) We can check this as well while doing the update

On 2019-07-31 09:01, stpr-unb wrote:

Thank you. A follow up question - I should be able to drop it into the LIBLINEAR directory and compile the mex file as I would for the standard LIBLINEAR correct?

-- You are receiving this because you commented. Reply to this email directly, view it on GitHub [1], or mute the thread [2]. [ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "https://github.com/cjlin1/liblinear/issues/54?email_source=notifications\u0026email_token=ABI3BHQE27ZHNSWRAZ52FL3QCGZNVA5CNFSM4IFO3ZN2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3HXQMA#issuecomment-516913200", "url": "https://github.com/cjlin1/liblinear/issues/54?email_source=notifications\u0026email_token=ABI3BHQE27ZHNSWRAZ52FL3QCGZNVA5CNFSM4IFO3ZN2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3HXQMA#issuecomment-516913200", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } } ]

Links:

[1] https://github.com/cjlin1/liblinear/issues/54?email_source=notifications&email_token=ABI3BHQE27ZHNSWRAZ52FL3QCGZNVA5CNFSM4IFO3ZN2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3HXQMA#issuecomment-516913200 [2] https://github.com/notifications/unsubscribe-auth/ABI3BHTIEPEOCRZYLFFF7XDQCGZNVANCNFSM4IFO3ZNQ

[stallam-unb]

Thanks a lot, I would appreciate it if you can look into the MATLAB generating code as well.

[ZYQue]

The updated ovo is available now, and it supports MATLAB. Please see the new release.