pilooch

30 ms·

301.
by pilooch 13y ago
Interested in the reasons behind the 'no brainer' move. We're in France, and the regime for startups, with innovation tax cuts and other security nets is pretty favorable. The engineers are pretty good too!
302.
by pilooch 13y ago
Check this out maybe, http://www.reddit.com/r/MachineLearning/comments/1dzoim/a_su...
303.
by pilooch 13y ago
With others, I recently played around with a custom built dataset, using unsupervised machine learning to cluster the subreddits, based on content. You can play with it here: http://demo.xplr.com/xplr/umbreddit/ The system can be queried:
304.
by pilooch 13y ago
I'm pretty sure this would be useful to me, but at install it does ask for 'access your data on all website'. Now, what does this mean exactly ? I do understand this is very much probably not this extension's fault, but Chrome being ver
305.
by pilooch 13y ago
Good job. But I need to make the note that the title is confusing to NLP/ML practitioners. 'topics' usually refer to clusters as captured by so-called 'topic models' [1], the output of an unsupervised learning method, usually a variant of L
306.
by pilooch 13y ago
It seems the relationship between the brain and deep learning has evolved in such a way that the later can help with insights into how the former works. In this regard, I thought I would mention the extraordinary simple and elegant talk by
307.
A new detector has a clever way of detecting dark matter(arstechnica.com)
2 points by pilooch 13y ago|0 comments
308.
Hi-res closeup of Curiosity (February 2013)(360cities.net)
1 points by pilooch 13y ago|0 comments
309.
by pilooch 13y ago
Sure, using TC hash datastore over millions (tenth of, not hundreds or billions) of entries, each being a compressed protobuffer. The cost of each write grows exponentially after some time. We've played with parameters, buckets size and num
310.
by pilooch 13y ago
Any hint at how http://symas.com/mdb/ behaves ? Been using Tokyo Cabinet for a long time now, and have repeatedly hit similar hangs. Shopping for a future datastore of this sort!
311.
The Brain: Our Food-Traffic Controller(nytimes.com)
1 points by pilooch 13y ago|0 comments
312.
by pilooch 13y ago
Amazing work, thanks for sharing. Any hint on how it behaves on big datasets ? Over time I've found that scaling up to real world (and industry) datasets requires going custom (or distributed, a-la-mr / hadoop).
313.
by pilooch 13y ago
Sure, I started as a researcher 14 years ago, then drifted to what I thought was a sweet spot then, half-research / half-programmer. I say 'sweet spot' because many applications did require both the academic and the applied background at th
314.
by pilooch 13y ago
Very interesting, but as a daily practitioner I am skeptical. First, this is a lot of code! As a C++ machine learning programmer, I am impressed as I know the pain (someone explains why, see comment https://news.ycombinator.com/item?id=5
315.
Sivacycle Atom - Power on Bike(kickstarter.com)
2 points by pilooch 13y ago|0 comments
316.
by pilooch 13y ago
I've upgrade a couple of days ago. Skype works fine, I'm using version 4.1.0.20, not sure if it was upgraded during the process. Chromium works fine.
317.
by pilooch 14y ago
easily gamed, type xplr, second link.
318.
by pilooch 14y ago
For the sake of realism only: any basic supervised machine learning algorithm does that with proper labeled dataset. I have built many. I can assure you this is so classic that Google has it, that many other companies have it. And there are
319.
by pilooch 14y ago
How frustrating these days to hit a science paywall. The original article would be worth reading.
320.
by pilooch 14y ago
I find Clang to be outdated in most Linux distributions. Asking developers here, are you compiling your own clang/libc++ to benefit from the continuous stream of improvements ?
321.
by pilooch 14y ago
Working with people who use org-mode to schedule and sort their daily professional activities with it when you're not is... quite an interesting experience. I remember this guy popping up every time on the channel right at 3:00pm sharp, as
322.
Show HN: real-time news dataviz with unsupervised machine learning(xplr.com)
3 points by pilooch 14y ago|0 comments
323.
by pilooch 14y ago
Recently a big engineering firm (let's not harm them by revealing their name :) ), called on me. A query on our system takes 7 days they said, when it succeeds... These guys deal with up to 4Tb a day. With proper multi-core map-reduce and o
324.
Europeans Reach Deal With Google on Searches(nytimes.com)
3 points by pilooch 14y ago|0 comments
325.
by pilooch 14y ago
Let me second that. Remain cold turkey for a little while, and grab a side project for consuming your energy on evenings. Some code you'd like to try on some new device, anything. And get medical attention, ask your friends for a loan if ne
326.
by pilooch 14y ago
This is a problem well-known in Web search as 'position bias': top results get more clicks and therefore get 'upvoted' by machine learning algorithms. There are theoretically sound solutions. Typically, I recommend [1]. Very briefly the sol
327.
by pilooch 14y ago
Sure, https://xplr.com/ we're seed funded, early stage, early customers.
328.
by pilooch 14y ago
I feel like adding another anecdotal fact: we're a startup in Toulouse and NY, and we love Toulouse, great city for both daily life and business (think Airbus etc...).
329.
by pilooch 14y ago
"The military are already in charge of subway stations and doing surveillance in front of important buildings (like Notre Dame). There are only going to be more and more of them. If the situation gets worse I expect the "couvre feu" to be p
330.
by pilooch 14y ago
Sure. Low (or even no) taxes on yearly income if you locate your startup well (e.g. put it in an area that gets subventions or tax cuts, checkout so-called ZFU aka Zone Franche Urbaine). If you're innovating (which is, really, an administra

More ›