idunning

8 ms·

31.
by idunning 11y ago
"Try it and see" will be pretty definitive (assuming you can implement it correctly, which is nontrivial), but it has scaling problems. I'd argue thats exactly what this talk is about: the presenter claims to have a new metho
32.
by idunning 11y ago
Agree that everything must be contextualized. My point is, you saw this presentation and liked it because it seems like a good idea, and the performance is apparently good relative to something else. For an engineer (or a theoretician) to
33.
by idunning 11y ago
Sure, but thats exactly my point. Given a set of well-stated limitations (like, cannot use more than N^x bytes of storage for an instance of size N, using the same standard library of data structures etc) we can make meaningful comparisons.
34.
by idunning 11y ago
Allowing precomputation seems a little suspect, or at least difficult to incorporate into a comparison. I could offline precompute all pairs shortest path, then I'll be 10^6 faster than Astar, for example. Where do you draw the line? I
35.
by idunning 11y ago
Not sure they are on HN, but I know a few companies at least are using the JuliaOpt[1]/ JuMP[2] stack. [1]: http://juliaopt.org [2]: https://github.com/JuliaOpt/JuMP.jl
36.
by idunning 11y ago
Ouch, "just sad"... would be interested to hear more! Even if you can't attend, anything you could write up as either a blog post, a Github issue, anything, would be helpful for getting some discussions going around it at the
37.
by idunning 11y ago
No, I don't think so. I don't think there is much that is going to change with modules in the future, or at least I'm unaware of any open issues about it. What did you have in mind? The hackathon on Wednesday will be a partic
38.
by idunning 11y ago
Not definite plans, but serious thought is being given to having it in Europe next year if all the necessary bits can come together.
39.
by idunning 11y ago
If anyone has any questions about this post or the conference, fire away! Last year's first conference was a big success in my opinion. Videos are available at http://juliacon.org/2014/ , although the speakers were
40.
JuliaCon 2015 schedule posted(juliacon.org)
2 points by idunning 11y ago|0 comments
41.
by idunning 11y ago
I'd say that the actual news is that the Julia founders have found a way to make a living through consulting, enabling them to keep working on an free (in both senses) software. They've also managed to scale to bring on some large
42.
by idunning 11y ago
It is an interesting point. In my usage, I don't view it as a one-or-the-other proposition. I use Julia, so if I'm doing some exploratory work and experimentation, or making something to present results, then the IJulia notebook i
43.
by idunning 11y ago
Great stuff from Norvig, as per usual. I find it strange that the next step after enumeration is a heuristic approach, something that comes up a lot in "solve a TSP" posts - there are many ways to solve TSP to provable optimality
44.
by idunning 11y ago
The really unfortunate thing about that post is it completely fails to understand the essential part of solving the TSP this way, which are the lazily-added subtour elimination constraints. I don't think the blog author at Forio really
45.
by idunning 11y ago
I wrote the solver code used here as an example for the JuMP modeling package for Julia, which is available at https://github.com/JuliaOpt/JuMP.jl . The specific TSP example code is available at https://githu
46.
by idunning 11y ago
Amazing, I've tried to do this for Wipeout 3 in the past (extract the tracks) and was defeated - this makes it seem fairly easy!
47.
by idunning 12y ago
CI against package ecosystems is a really great idea. We do it for Julia too [1], and it can identify some really subtle issues that would otherwise take longer to become apparent. http://pkg.julialang.org/pulse.html
48.
by idunning 12y ago
Original content is at http://nikhilbuduma.com/2015/03/10/the-curse-of-dimensionali...
49.
by idunning 12y ago
I don't think its the same, but there is a high-quality implementation in Julia that is registered: https://github.com/dfdx/Boltzmann.jl
50.
by idunning 12y ago
Some techniques, e.g. random forest, give variable importance indicators for free. If you can test it out, give it a go - don't have to use the random forest as the final model.
51.
by idunning 12y ago
If you are disciplined, and separate data into training and testing sets, you can try as many models as you want without fear of overfitting. Indeed, optimizing over the parameters of a model on the training set is essential (pruning parame
52.
by idunning 12y ago
The author mentions that one of his goals was to focus on a smaller set of functionality and make it simple and high-performance, but I've got to put a shoutout to Mocha.jl here [1]. It is essentially Julia's answer to the Caffe d
53.
by idunning 12y ago
PuLP, apart from being in Python and this being in Ruby, is definitely a more general tool as it can connect to a wide variety of (MI)LP solvers. Heres a link to the latest version of PuLP: https://pypi.python.org/pypi/
54.
by idunning 12y ago
An alternative is JuMP [1], a package for Julia. You can model optimization problems with linear, quadratic, and general nonlinear constraints/objective and send them to a variety of open-source and commercial solvers. JuMP, these solv
55.
by idunning 12y ago
In my opinion, as a contributor to Julia and someone who teaches machine learning with R - start with R. Things will "just work" for the most part and you won't have to worry about whether your packages will work while you ar
56.
by idunning 12y ago
How is that dependency graph plotted in the README?
57.
by idunning 12y ago
Are you talking about compiling R from source, including all dependencies and high-performance BLAS, LAPACK, FFT, ARPACK, etc. libraries? If not, you are comparing apples to oranges.
58.
by idunning 12y ago
I feel like people are perhaps a bit overly negative about the state of Julia's packages. Compared to other early-stage languages, I'd say our package ecosystem is vibrant and full of fantastic packages. Naturally perhaps they are
59.
by idunning 12y ago
Countering your counter-point: this seems like the most pessimistic comment about doing a PhD I've ever seen. There are many types of department, many fields, that don't operate like the type of setup you are describing. I'd
60.
by idunning 12y ago
I should point out the main Julia package listing is at http://pkg.julialang.org/ , this is the "executive summary" page. At the bottom of this "Pulse" page are test results - we run the tests for every p

More ›