manthideaal

7 ms·

31.
by manthideaal 7y ago
My understanding from the discussion part, section 5 of the linked paper, is that the GAN could be modified so that the relative power of the discriminator and generator is fine tunned in order to generate hard to detect images, by giving
32.
by manthideaal 7y ago
From the paper: (1) We show that when the correct steps are taken, classifiers are indeed robust to common operations such as JPEG compression, blurring, and resizing. (2) When using Photoshop like methods the detector performs at chance (i
33.
by manthideaal 7y ago
Using Kolmogorov Complexity and then using PCA is not valid since you are approximating a solution and Kolmogorov Complexity is for exact solutions. Anyway perhaps there is, or should be defined, a signal/noise Kolmogorov Complexity me
34.
by manthideaal 7y ago
You (me) must read something about backscattering, a quick suggestion is: https://www.sciencedirect.com/topics/computer-science/backsc...
35.
by manthideaal 7y ago
Just a reflection about this: Could a modification of this backscattering technique be used so that hundred of this devices use the same channel to transmit a set?, I am thinking about collecting the information of hundred of devices each
36.
by manthideaal 7y ago
In the paper there are some bounds about the number of states in the automata as a function of the length of the input. So one could limit the length of the input when using back references to bound the complexity of the algorithm. They hav
37.
by manthideaal 7y ago
See theses two answers: https://stackoverflow.com/questions/39396491/is-there-a-way-... https://stackoverflow.com/questions/39255471/what-is-the-poi...
38.
by manthideaal 7y ago
if you call a function in the form module:function-name then erlang detect and use the latest compiled version of the module, if you use function-name without the module then the old version is used. Is advised to use the module:function-n
39.
by manthideaal 7y ago
In erlang the code is compiled to an intermmediate representation for a virtual machine, BEAM (erlang virtual machine), usually the functions you define are very short and in this case the BEAM code is sent over the network, so the univers
40.
by manthideaal 7y ago
You can find more information about erlang in this Ph.D Thesis (1), on page 86 is an example of a universal server, explained with detail. https://erlang.org/download/armstrong_thesis_2003.pdf
41.
by manthideaal 7y ago
Perhaps to protect against ReDoS the client should use an extended finite automata (1). https://www.arl.wustl.edu/~pcrowley/a25-becchi.pdf (1) Extending Finite Automata to Efficiently Match Perl-Compatible Regular Ex
42.
by manthideaal 7y ago
I agree with you, the phrase should be preceded with: Caution, this is a math joke, don't lose sleep trying to grok it.
43.
by manthideaal 7y ago
I wonder what's the point of using such a phrase, it doesn't help you to grok the concept of monad. It can help you to know that someone has given a new definition to sound cute, shame on them. By the way I admire MacLane as a ma
44.
by manthideaal 7y ago
The definition of a monoid here is not the usual definition, is a new definition for the special case of a strict category as defined MacLane's Book Categories for the Working Mathematicean. Since you maybe thinking about composition
45.
by manthideaal 7y ago
I am not able to wait for a video to explain something, I prefer to read books. I don't know what is good about this teacher. Linear Algebra Done Right is a good book. I am not able to appreciate Gilbert Strang style. But algebra is
46.
by manthideaal 7y ago
It seems you learned from the decline of Perl, so you have the experience of a warrior. But Rhombus represents the hope of a better future for Racket, so another step in this fight should be to embrace the future and learn from the past, I
47.
by manthideaal 7y ago
Thanks, nice. bytes.fromhex is really reading bytes from a hex string, so perhaps a better name should be bytes.fromhexstring, but since python names are short I understand the trade-off. Also, my first impression is that hex should be call
48.
by manthideaal 7y ago
Why?, not a native English speaker, I looked up doge and it has no meaning, any tips for other meaning?
49.
by manthideaal 7y ago
python, decoding and encoding the message a="53414e544120414e442048495320574f524b53484f50" "".join([chr(int(a[i:(i+2)],16)) for i in range(0,len(a),2)]) => 'SANTA AND HIS WORKSHOP' a = 'SANTA AND HIS WO
50.
by manthideaal 7y ago
About the work you cite, I think that double descent is simply because the extra number of parameters (low bias) used produces a large variance when the input data is small, but as more data is introduced the extra parameters don't pla
51.
by manthideaal 7y ago
The critique is about the importance of priors in BNN. In my humble understanding of Bayesian reasoning the argument to defend any prior is that with enought data the learning method converges to the real distribution, so if the result o
52.
by manthideaal 7y ago
The example of misuse of the CLT illustrates that a mixture of normals is not a normal distribution. For example the height of people is a mixture of two normals, one for men and the other for women, this mixture has two modes and so it