nestorD

24 ms·

271.
by nestorD 5y ago
You can but it is both costly and a hack (the resulting embedding will not be as good as the one you would have gotten restarting from scratch). So I would not recommend using it in an inference pipeline.
272.
by nestorD 5y ago
The memorized set goes away fairly easily. However, profesional mnemotechnicians who do several demonstrations a day will sometime take the time to 'destroy' the pictures in between demonstrations (setting them mentally on fire fo
273.
by nestorD 5y ago
I would be interested in a reference because this goes directly against my experience (I have used this method to recall lists of random words given by a group of people) and the papers I read.
274.
by nestorD 5y ago
You can (and I do) use the fact that it is a 3D space to store further information (such as ordering and relationship between items). In CS terms, a memory palace is usually used as a linked list while a peg system would be a hash-table but
275.
by nestorD 5y ago
Note that I have seen a chicken coop devasted by a dog in exactly the same way. It was not a pretty sight.
276.
by nestorD 5y ago
F# comes to mind. It uses indentation instead of braces and has a nice ecosystem.
277.
by nestorD 5y ago
That (and overall the unability to take correlation between error into account) and the fact that intervals propagated are upper bounds for some things like the division.
278.
by nestorD 5y ago
You should be okay with a sum but interval arithmetic can blow up in a way that is not correlated with the numerical accuracy of the algorithm. Newton's method, for example, is strongly convergent but produces diverging intervals.
279.
by nestorD 5y ago
Bioinformaticists have been able to do that with traditional algorithms for years (dynamic programming gets you a long way to compute an edit distance for example). It is probably the first thing that was done once the COVID-19 genome was m
280.
by nestorD 5y ago
Interestingly, one of the fathers of weather forecasting wanted to predict the likelyhood of war using similar mathematical tools: https://nautil.us/issue/70/variables/cloudy-with-a-chance-of...
281.
by nestorD 5y ago
APOPO (the association behind this initiative) lets you "adopt" a rat: in exchange for money they give you regular information on what and how it is doing. This sounded like a great gift so I did that for a friend but, sadly, pass
282.
by nestorD 5y ago
For me the good reason to use gaussian regression is the fact that you get an uncertainty on the output. The big downside is that it takes expert knowledge (to design a proper kernel) and a solid implementation (to avoid the various numeric
283.
by nestorD 5y ago
On the subject of addiction I highly recommend this comic on the rat park experiment: http://utw10426.utweb.utexas.edu/quest/Q3/ratpark.html It starts with the famous experiments where rats would prefer drugs to f
284.
by nestorD 5y ago
> She attributed the mess to one Medici in particular, Alessandro Medici, a ruler of Florence, whose assassinated corpse had apparently been buried in the tomb without being properly eviscerated. Over the centuries, he seeped into Michel
285.
by nestorD 5y ago
As said in another comment, I am very happy with Xournal++. It let you add pictures (I draw my signature once in Photopea and saved the file) and text (useful to fill forms).
286.
by nestorD 5y ago
I am very happy with Xournal++ (which can also be used to add text to the pdf in order to fill a form for example).
287.
by nestorD 5y ago
That's very strictly forbiden. The fastest-way-to-get-fired kind of forbiden.
288.
by nestorD 5y ago
I would add that I had the same problem as a foreigner moving to the US. The number of services (both public and private) that require you to have a US phone number was both surprising and impressive.
289.
by nestorD 5y ago
I wrote pandas2numpy [0] it helps you turn a pandas dataframe into one (or two if you have categorical variables) numpy tensors ready to be pushed into a deep learning framework. It can deal with continuous data, categorical data, normalisa
290.
by nestorD 5y ago
my problem with that is that some speakers pick distracting shirts (the worst offenders have text on their shirt which I cannot help but try to read instead of listening to them).
291.
by nestorD 5y ago
`Learn Rust With Entirely Too Many Linked Lists`[0] is a great tutorial on the subject. [0]: https://rust-unofficial.github.io/too-many-lists/
292.
by nestorD 5y ago
I discovered computer science in preparatory classes (a French thing that comes before engineering school, you can think of it as the first years of college) in France where it was the default programming language (I believe they now start
293.
by nestorD 5y ago
> I find it fascinating how, 25 years on, many of the most popular languages are still slowly encorporating core features of functional languages like OCaml Yes! Ocaml was my first programming language and I have been baffled at how lang
294.
by nestorD 5y ago
> The doc contains cheatsheets such as "Gleam for X users" where X is currently Erlang, Elixir, Rust and Python. I think this is a fantastic idea and could be adopted by other language communities! The Rust ndarray library as a
295.
by nestorD 5y ago
There is such a stone marker in my parent's garden (their house is definitely fully in French teritory)! To be fair they are living so close to the frontier that my cellphone picks up the Belgian cell network in their house...
296.
by nestorD 5y ago
From memory you don't need size to be a prime, what you need is that it is coprime with your step. Thus, the easiest solution is to take a step-size that is prime and different from you size (that works for any size). Given the index
297.
by nestorD 5y ago
The links appears to now be youtube.
298.
by nestorD 5y ago
The model takes pictures, encoding the material structure, as inputs and not photographs. Seing the pictures in the article, the model produces very coarse results from very simple inputs.
299.
by nestorD 6y ago
I personnaly went from Ocaml to F# because it had much better tooling (IDE support) and a larger libraries ecosystem (access to all .Net libraries).
300.
by nestorD 6y ago
One big caveat is that they use a parallel algorithm on a 44 cores machine to compete with a single GPU. I can train a neural network on the GPU in my laptop but it has only 8 cores.

More ›