<div dir="ltr"><div>Hi Ryan,</div><div><br></div>Thanks for your advice. I would follow your suggestion and see what I can find.<div><br></div><div>Best,</div><div>Zhaoduo</div></div><div class="gmail_extra"><br><div class="gmail_quote">2016-03-18 13:08 GMT+08:00 Ryan Curtin <span dir="ltr"><<a href="mailto:ryan@ratml.org" target="_blank">ryan@ratml.org</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Thu, Mar 17, 2016 at 10:21:33PM +0800, Zhaoduo WEN wrote:<br>
> Hi Ryan,<br>
><br>
> I want to figure out one thing. Does this project intend to expand the<br>
> existing mlpack_cf? Could I write a new algorithm that is different from CF?<br>
><br>
> I've read the API of CF, though not all of them. I find that it's not easy<br>
> to fit with the existing other factorizers and algorithms for CF, as the<br>
> key component of SLIM is learning W matrix from A. However, the CF<br>
> algorithm doesn't need this. Do you have any ideas of fitting with the existing<br>
> other factorizers and algorithms for CF? I need some inspiration. Thanks.<br>
<br>
</span>Hi Zhaoduo,<br>
<br>
I glanced at the paper and I see what you mean; instead of decomposing<br>
the data matrix into two low-rank matrices W * H, the SLIM algorithm<br>
instead simply learns a large sparse matrix W.<br>
<br>
I think it would be very useful to allow the SLIM algorithm to fit<br>
inside the CF code, but clearly the code that is currently in CF will<br>
not work and it will need to be refactored to handle more models than<br>
just a low-rank decomposition.<br>
<br>
Probably the right way to go about this will be to refactor the CF class<br>
so that the logic for the model is contained in the FactorizerType or<br>
something else, not in the CF class itself. So it may be necessary to,<br>
e.g., define a ModelType class associated with each FactorizerType which<br>
contains the logic for applying the model to make a prediction.<br>
<br>
I haven't thought through all the details there, but I think that should<br>
be a good direction for how to resolve this problem. Let me know what<br>
you think.<br>
<br>
Thanks,<br>
<br>
Ryan<br>
<span class="HOEnZb"><font color="#888888"><br>
--<br>
Ryan Curtin | "I know... but I really liked those ones."<br>
<a href="mailto:ryan@ratml.org">ryan@ratml.org</a> | - Vincent<br>
</font></span></blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature"><div dir="ltr"><div><div dir="ltr">Best Regards,</div><div dir="ltr"><br><div>Zhaoduo</div></div></div></div></div>
</div>