liblinear solves the objective
1/2 * R + C * L
where R and L are the regularization and loss terms.
Our SGD implementation solves the objective
alpha/2 * R + 1/ n_samples * L
To go from the former to the latter, we can divide by C and by n_samples
(which doesn't change the solution). So alpha = 1 / (C * n_samples).
Mathieu
On Fri, Dec 27, 2013 at 8:18 PM, Andy <[email protected]> wrote:
> On 12/17/2013 11:39 AM, Joel Nothman wrote:
> > I think alpha = 1/2C
> >
> I think alpha = n_samples / C (not sure about the 2)
>
>
> ------------------------------------------------------------------------------
> Rapidly troubleshoot problems before they affect your business. Most IT
> organizations don't have a clear picture of how application performance
> affects their revenue. With AppDynamics, you get 100% visibility into your
> Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics
> Pro!
> http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
> _______________________________________________
> Scikit-learn-general mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>
------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
