groar

11 ms·

91.
Mocha.jl: Deep Learning for Julia(devblogs.nvidia.com)
66 points by groar 11y ago|2 comments
92.
Mocha.jl: Deep Learning for Julia(devblogs.nvidia.com)
3 points by groar 11y ago|0 comments
93.
On the Expressive Power of Deep Learning: A Tensor Analysis(arxiv.org)
59 points by groar 11y ago|3 comments
94.
Deepomatic Grabs $1.4M for Its Deep Learning Visual Search Technology(techcrunch.com)
13 points by groar 11y ago|0 comments
95.
Predicting image ads click rate using convolutional neural networks(arxiv.org)
1 points by groar 11y ago|0 comments
96.
Scalable Bayesian Optimization Using Deep Neural Networks(arxiv.org)
63 points by groar 11y ago|3 comments
97.
Deep learning pyschology: Rorschach images seen by classifiers(medium.com)
2 points by groar 11y ago|0 comments
98.
StochasticNet: Forming Deep Neural Networks via Stochastic Connectivity(arxiv.org)
1 points by groar 11y ago|0 comments
99.
by groar 11y ago
For those who never heard about Tierra. I bumped into it today, and it reminded me of the times I was fascinated by this experiment and was playing with similar ideas. Old school.
100.
Tierra experiment: good old times evolution simulator(en.wikipedia.org)
1 points by groar 11y ago|1 comments
101.
Neural Transformation Machine: Sequence-To-Sequence Learning(arxiv.org)
49 points by groar 11y ago|1 comments
102.
Gaussian Process School Notes(gpss.cc)
36 points by groar 11y ago|1 comments
103.
by groar 11y ago
Yeah but it's still misleading as we are talking about approximating continuous functions here, not any function. Those examples are not clearly computable, or even just continuous..
104.
by groar 11y ago
For those who want to dive further in automata theory, transducers and the non-commutative algebraic view of the theory, the following one by J. Sakarovitch is a fantastic book : http://www.cambridge.org/uk/catalogue&#x
105.
ImageNet 2015 Challenge registration is up(image-net.org)
1 points by groar 11y ago|0 comments
106.
by groar 11y ago
Clearly, when I think about the hardest thing I ever coded, I have the following story in mind. Back in 2002 I was writing a floppy disk driver for the little OS we were writing with a friend. It turned out finding anything else than very s
107.
by groar 11y ago
I attended the 2011 edition, and this was one of the best experiences I had during my PhD. Lots of brilliant people, amazing lectures, blue grass music, a rafting session, great beers and so much more. You have to go.
108.
by groar 11y ago
There are many reasons for the name: apart from being one of France's emblems, one of the people who initiated the project is Thierry Coquand and the original type system underlying it is called Calculus Of Construction (CoC). But I kn
109.
by groar 11y ago
The problem with memory models is not some much verifying it in a mechanised way, but inventing something suitable at all. Agreed. I'm aware of some work done a few years ago by people working on weak memory models and extendin
110.
by groar 11y ago
What you probably mean is something like nice abstract accounts of memory models for C that at the same time capture all the optimisations modern C-compilers want to do, while still being implementable on existing CPUs. That is indee
111.
by groar 11y ago
In a rich type system, a program is the proof of its own specification. Which is usually as interesting as the program itself.