pilooch

26 ms·

271.
by pilooch 13y ago
A very good source of information, and tutorials with source code is http://deeplearning.net/tutorial/contents.html
272.
by pilooch 13y ago
hyperparameter search is a solution. See http://jmlr.org/papers/v13/bergstra12a.html and http://jaberg.github.io/hyperopt/
273.
NIPS Deep Learning Workshop has Mark Zuckerberg on panel(sites.google.com)
1 points by pilooch 13y ago|0 comments
274.
Build Real-time Big Data Applications on Apache HBase.(kiji.org)
3 points by pilooch 13y ago|0 comments
275.
by pilooch 13y ago
put it on fdroid
276.
by pilooch 13y ago
Hopefully self driving car are on the way ;)
277.
by pilooch 13y ago
interesting but why is this a remote service that needs to parse in-browser content? if filtering is keyword based (as opposed to ml based), why not filter locally instead ?
278.
by pilooch 13y ago
FYI I use Solr cloud for that purpose.
279.
by pilooch 13y ago
Interestingly, Linux is moving to win on several fronts. Economy and debt crisis, in addition to rising Ubuntu-like UI (easy to use, LibreOffice included, etc...) increase the benefits for governments and agencies. That's one front. On
280.
by pilooch 13y ago
I love rien, I am a Parisien ;)
281.
When Startups Seek Patent Trolls as Saviors, the System Has Failed(eff.org)
4 points by pilooch 13y ago|0 comments
282.
by pilooch 13y ago
The title is a bit confusing as probabilistic programming is a research field itself that the book seems to not touch upon. See http://probabilistic-programming.org/wiki/Home
283.
Ask HN: why Google Plus has no built-in RSS?
1 points by pilooch 13y ago|0 comments
284.
by pilooch 13y ago
yeah right, and my bill looks like this: http://i.imgur.com/WxZCFrH.png Thanks to Google maths I presume!!
285.
Duke Nukem 3D code review(fabiensanglard.net)
3 points by pilooch 13y ago|0 comments
286.
Hacking machine learning with... machine learning(arxiv.org)
1 points by pilooch 13y ago|0 comments
287.
by pilooch 13y ago
hi, thanks for your interest, the source code is here: https://sourceforge.net/p/seeks/code/ci/experimental/tree/
288.
Ask HN: I've started an OSS decentralized search engine a while ago / need help
7 points by pilooch 13y ago|3 comments
289.
by pilooch 13y ago
Right, one way is to share one instance with someone you trust (and who runs a linux server), and to use it remotely.
290.
by pilooch 13y ago
Probably none yet, unfortunately. Seeks is usable because it relies on Google or any other engine in the background, when no results are available among peers. Yacy is more advanced, uses a DHT, but does not really work for daily use, the a
291.
by pilooch 13y ago
DDG is nice initiative. The problem is that it does not solve any problem related to privacy because it is based on a matter of trust. Distributed search, similar to bittorrent, DHT-based designs and the like are notoriously difficult. I&#x
292.
by pilooch 13y ago
So, actually one I've discovered, and tried with great success, is Mumble, http://mumble.sourceforge.net/
293.
by pilooch 13y ago
Someone here said 'Champagne holder' ;)
294.
by pilooch 13y ago
many thanks! It may well be the one.
295.
by pilooch 13y ago
right, we already use an IRC server over VPN. Definitely the best for chat and logs I agree!
296.
Ask HN: serious alternative to Skype on Linux/OSX/Windows and smartphones?
15 points by pilooch 13y ago|8 comments
297.
Ask HN: has Twitter RSS / Atom search been cut?
2 points by pilooch 13y ago|0 comments
298.
Campaign Finance Linker(github.com)
1 points by pilooch 13y ago|0 comments
299.
by pilooch 13y ago
haha, everyone would like to kill URSAFF, but they are your life after your biological CPU has turned into a good old Pentium :) And yes Limousin, may not be your startup hub of choice, but after all, it is probably cheaper than everywhere
300.
by pilooch 13y ago
For the past month, I've been checking on G+ before Twitter and other media or sources. Here is how I was dragged into it. I had never used Facebook, and I had never used Google+. I guess I was too busy with Emacs. Then recently I had to hi

More ›