Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
nestorD
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
55 ms
·
541.
▲
by
nestorD
8y ago
Same here, it seems that amazon has improved its protection and changed its ebook format to harden it against DRM removal.
542.
▲
by
nestorD
8y ago
The benchmarks seem to be limited to small functions, my big question would be how does prioritized grammar enumeration scales to large problems (many variables) compared to Grammatical evolution ?
543.
▲
by
nestorD
8y ago
Watching the following TED talk, I was surprised to discover that gas compagny such as Total are actually promoting renewable energies because they currently need to be complemented and gas is the main solution used for that purpose. https
544.
▲
Symbolic Regression, Genetic Programming or If Kepler Had R
(blog.ephorie.de)
7 points
by
nestorD
8y ago
|
2 comments
545.
▲
by
nestorD
8y ago
As commented by others, this is the kind of functions for which you want to use a numerically stable algorithm. For that you can : - sort the numbers (clearly not the most cost effective solution but easy), - use a high precision accumulato
546.
▲
by
nestorD
8y ago
My feeling is that it is not about improvement but democratisation : Ocaml has been around since 1996 and its type inference is stellar. I was shocked when I first touched C and Java after two years of Ocaml.
547.
▲
by
nestorD
8y ago
F# might be a good fit: it is mature, has modern tooling and, in my experience, people tend to get productive quickly.
548.
▲
by
nestorD
8y ago
lulu.com is such a company (not affiliated but I have friend that have used it).
549.
▲
by
nestorD
8y ago
A fibonacci pseudo random number generator would not be too hard to keep in mind (two integers modulo a fixed number) and might give you ok results if you choose your parameter well and avoid using the weakest bits as the generated boolean
550.
▲
by
nestorD
8y ago
Currently in the second year of a happy PhD in France. Looking at other PhD-students around me I would say that those that have few results so far are indeed stressed, but clearly not to those levels, and that the main criteria for their we
551.
▲
by
nestorD
8y ago
What Every Computer Scientist Should Know About Floating-Point Arithmetic is great because it makes people aware of problems with the current system. Adding UNUMs would be suggesting the devil you don't know over the devil you know.
552.
▲
by
nestorD
8y ago
There is a xkcd package to fulfill this need : https://stackoverflow.com/questions/12675147/how-can-we-make...
553.
▲
Putting the brakes on high-frequency trading with physics
(hackaday.com)
2 points
by
nestorD
8y ago
|
0 comments
554.
▲
by
nestorD
8y ago
> We find that, while -O2 has a significant impact relative to -O1, the performance impact of -O3 over -O2 optimizations is indistinguishable from random noise.
555.
▲
by
nestorD
8y ago
Not equal apparently : https://youtu.be/mICxKmCjF-4
556.
▲
by
nestorD
8y ago
There is a "filter similar message" option available (it is hidden with the answer /transfer / print / etc options).
557.
▲
by
nestorD
8y ago
And that's just meaning. I found that preserving the style can be an equally hard thing to do from a language to another.
558.
▲
by
nestorD
8y ago
A trick with the memory palace is to use place you know well such as your bedroom. Then, with practice, you will learn how to build image that are vivid to your mind and easier to remember (it is perfectly normal if at first you build image
559.
▲
by
nestorD
8y ago
I recall reading that there is an, incorrect, proof that would match the kind of proof we expect from Fermat and thus is believed to be the one he had in mind. However I was unable to find it. Instead I found a discussion on how likely it i
560.
▲
by
nestorD
8y ago
The absence of national ID just means that a substitute, tat was not designed for this purpose (such as a driver's license), is used instead. This introduces its own set of problems as illustrated by this video on voter registration in
561.
▲
by
nestorD
8y ago
Just discovering the game but the imperfect information and convergent nature of the game looks like a good fit for a monte-carlo based algorithm.
562.
▲
by
nestorD
8y ago
The documentation could clearly be prettier but its strength lies in its standardization : I can get method arguments and examples for all functions of all packages in the same format and with a single keystroke.
563.
▲
by
nestorD
8y ago
I would put it after R help function and, even better, F1 key to get the full documentation of any thing in Rstudio.
564.
▲
Random Search for Hyper-Parameter Optimization (2012) [pdf]
(jmlr.org)
1 points
by
nestorD
8y ago
|
0 comments
565.
▲
by
nestorD
8y ago
Reading it I actually thought that it would be a perfect bedtime story for a child old enough. Thank you for your story.
566.
▲
by
nestorD
8y ago
On that subject (following a recommandation on HN) I started reading the Memoirs of the Conquistador Bernal Diaz del Castillo [0]. It is an account from a conquistator that was a member of the expedition that discovered the Aztec written in
567.
▲
by
nestorD
8y ago
I was told Coq but, after some checking, I can only find the B method indeed. However I did find a paper mixing both ([0]) : using Coq to prove the more complex proofs generated by atelier B. So it might have been both. [0] https:/&#x
568.
▲
by
nestorD
8y ago
The exemple I was given while studying the subject is that the sofware for the driverless metro line 14 in Paris has been proven in Coq.
569.
▲
by
nestorD
8y ago
Regarding data structure there is the Immer library : https://sinusoid.es/immer/#contents
570.
▲
by
nestorD
8y ago
Note that units are implemented in F# : https://fsharpforfunandprofit.com/posts/units-of-measure/
More ›