Acsor

general recent conversations

Discussion between Acsor and Yuval Fi…

Imported from a comment discussion on cs.stackexchange.com/que...

Acsor

Nov 19, 2021 14:06

It doesn't seem super hard either. Specifically, I worry I'm overlooking some critical piece of literature (textbook, article) that could be instrumental to the solution

Acsor

Nov 19, 2021 14:05

@Stef He he, if it was plain easy I wouldn't be asking for help :-)

Acsor

Nov 19, 2021 14:03

@Stef Yes, in that case $C =$ run-length encoding

Acsor

Nov 19, 2021 14:01

Ideally I'd like to determine an as large as possible logarithmically-compressible source $S$ for an algorithm $C$, not just giving the example of a string that can be logarithmically compressible. However, I don't need an extended theory of all sources where this is possible either

Acsor

Nov 19, 2021 13:59

Not sure if I'm misusing language here, but by source I mean any process producing those patterned, regular strings

Acsor

Nov 19, 2021 13:58

@Stef Yes, that's the point. In other words: fix a compression algorithm $C$ (extended Huffman, LZ77, LZ78, arithmetic encoding, whatever you want...) and determine what source $S$ (however narrow) lends itself to logarithmic compression

Acsor

Nov 19, 2021 13:41

Finally: yes, I'm willing to focus my attention to specific strings with a certain regularity, which is what I meant by "under what sources?" in my post title. I am new (and self-taught) to information theory so maybe I wasn't careful enough with the choice of words

Acsor

Nov 19, 2021 13:37

I'd also like to know how entropy relates to dictionary coders. Despite all the literature I've got my hands on, the two things are seldom related to each other, so I'm not able to make certain statements as "No, LZ77 can't work here because the entropy... such and such"

Acsor

Nov 19, 2021 13:35

>Shannon's source coding theorem applies to any compression scheme, not just character-by-character

If we consider a source that emits blocks of characters of length $k > 1$ and we assign a codeword to each of those blocks, and if $k$ is sufficiently long and the number of different blocks is low enough, shouldn't we able to achieve this type of compression?

Acsor

Nov 19, 2021 13:32

Hey, thanks for reaching out. Would you mind moving the conversation to the chat? Take your time, I don't want to interfere with your other duties!

Acsor

Nov 19, 2021 13:31

That was one of my concerns, indeed. Note, however, that I'm considering a logarithmic compression without regard to constant factors. Aside from that, it's somehow cumbersome (if not hard) to translate this problem from a function-based perspective into a rate-based one, which relates to the entropy of the source.

Acsor

Nov 19, 2021 13:31

That's, in fact, my reasoning for why the non-extended Huffman encoding cannot work. However, what if you don't correspond each symbol $x_i$ in the source string $X$ to another symbol in the output string $Y$? Shouldn't that change things?

Acsor

Nov 19, 2021 13:31

Also, the entropy relates to how many bits you need on average for the compression of an input string, meaning that it is possible to do better for a few, specific ones (provably so? that's what I'm trying to understand). It's okay for me to restrain the compression algorithm $C$ to a rather restricted scenario, even if unrealistic.

Computer Science

General discussion for cs.stackexchange.com

Acsor

Jun 11, 2021 14:11

To motivate my interest: I'm trying to determine plagiarism attempts in database schemas through ER or UML diagrams. The typology of database is mostly relational, which means I can also consider SQL source code. The interest in graph-similarity measures arises because one first attempt at addressing the problem comes from considering ER or UML schemas as graphs. However, I'm a mere undergraduate, and I'm not familiar (yet) with anything AI-related

Acsor

Jun 11, 2021 14:06

Hi, is there any graph-edit distance measure or algorithm that doesn't rely on NNs or statistical learning methods? I'm reading A survey of graph edit distance to find some, but getting some advice by someone who is knowledgeable wouldn't be bad

Acsor

Jun 10, 2021 16:00

I determined I cannot tackle that paper before I have taken a proper coures on ML/AI

Acsor

Jun 10, 2021 10:25

I'd like to determine if I have the necessary background to study that paper, and if not whether it's doable to acquire it in a reasonably short time span

Acsor

Jun 10, 2021 10:21

Hi, I'm trying to study the [Graph Kernel](http://jmlr.csail.mit.edu/papers/volume11/vishwanathan10a/vishwanathan10a.pdf) paper, which describes ways to compute similarity measures between edge-labeled graphs.

Unfortunately, I'm an undergraduate, and while most of the mathematical background is known to me, some of it isn't. Specifically, the authors talk about so called __feature maps__. What are they and where do I learn about them? It looks like it is a concept from ML/AI, but I'm not sure. They also mention Hilbert Spaces and RKHS, but in a less pervasive way

Acsor

Mar 13, 2021 10:18

Hi, thanks for answering. I saw it once in a Logic class while studying Zermelo-Fraenkel axiomatic theory, but that was in a rather different context. I may as well just ask directly the OP

Acsor

Mar 12, 2021 19:59

Hi, what does the $[\![1, n]\!]$ notation mean? I just found it in a question recently asked

Acsor

Feb 28, 2021 22:26

Oh, it appears he was establishing a mere ordering between terminals

Acsor

Feb 28, 2021 22:08

Hi, I'd like to answer this question, but what does OP mean by his a < b < c notation? It's something I had yet to see in the context of formal languages

Acsor

Aug 12, 2020 17:06

I like math. How do I act accordingly in deciding my CS career under these circumstances? I am still a mere "undergraduate", if that word applies in my country.

I ask this because sometimes CS looks really stimulating, and others *a lot more* discursive and boring. I cannot really guess what its true nature is, whether just one of the two options or both.

theory salon

theoretical computer science. highlight reel vzn1.wordpress.co...

Acsor

Aug 7, 2020 17:43

where M denotes a DPDA. Now the question is: isn't the second condition deductible from the first one, or am I simply missing something? Specifically, if the (q, a, x) triplet is such that to map to one element, if a = $\eps$ then obviously no other mapping exists for a different choice of a

Acsor

Aug 7, 2020 17:39

Acsor

Aug 7, 2020 17:39

Hi, I have a question that doesn't really warant its own post.

We are in context of (elementary) automata theory, specifically DPDA. My doubt concerns the following definition

Acsor

Aug 2, 2020 20:41

Thanks, looks really cool. I'm checking it out right now, but probably humans can still provide information machines cannot (e.g. university culture)? Still considering opening a question

Acsor

Aug 2, 2020 14:41

For clarification only: I'm a second-year CS "undergraduate" and would like to plan ahead my specialization degree. I'd like to know more about TCS, logic- and math-based stuff

Acsor

Aug 2, 2020 14:39

Hi, if I wanted to know more about available TCS master's degrees around Europe and opened a question in https://cstheory.stackexchange.com/ for doing so, are there any chances it would obtain serious replies or would it just merely get downvoted and closed for not respecting the Q&A policies?

Sorry to ask, but I have experience with past SE communities doing so

The Whiteboard

General Discussion for softwareengineering.stackexchange.com P...

Acsor

Oct 28, 2019 20:37

@JoelHarmon thanks for the clarification ;-) Now that I am at it though, what is meant by the intent itself of the pattern?

_The intent of the Builder design pattern is to separate the construction of a complex object from its representation._

I have always taken this as separating the construction of a complex object from its _interface_, not _representation_ (i.e. implementation). Is this just a terminology misunderstanding or am I not grokking the pattern?

Acsor

Oct 27, 2019 08:16

Can it refer to a class interface? I ask it because DP is at the same time a book with strict terminology and that same word is not defined

Acsor

Oct 27, 2019 08:15

What is meant by the technical term representation in GOF book Design Patterns? I'm referring to the use that is made e.g. in the description of the Builder pattern.

The h Bar

General chat for Physics SE (physics.stackexchange.com). For M...

Acsor

Aug 29, 2019 16:14

Thanks for the insight! :)

Acsor

Aug 29, 2019 16:14

Yes yes, that was my final point. I just misinterpreted the text

Acsor

Aug 29, 2019 16:10

Well it looks it all boils down to how I interpreted the statement, which is "The spring can be held compressed at L1 by an application of a constant force F1". Basic Newton 2nd law yields the answer

Acsor

Aug 29, 2019 16:05

What does "a spring that can be compressed 2.0 cm by a force of 270 N" mean then? 270N at the end? If it is, it sounded hugely ambigous to me :/

Acsor

Aug 29, 2019 15:55

Probably I have done the wrong assumption of equating the spring potential energy with the work from the constan force

Acsor

Aug 29, 2019 15:39

Someone is going to suggest to rely on the $F = kx$ equation, but then the question still applies. Why doesn't the reasoning above lead to the same result?

Acsor

Aug 29, 2019 15:31

Here I am, it's problem 29. I've done everything right, except determining the $k$ spring constant.

My reasoning was: if the spring can be compressed by $L1 = 2.0cm$, then during that process it accumulates $\frac{1}{2} k L1^2$ energy. If this energy is drawn from force $F1$ which acts along a distance $L1$, then we have the equation

$\frac{1}{2} k L1^2 = F1 L1$.

But as it turns out, something must be wrong in the reasoning. But what exactly?

Acsor

Aug 29, 2019 15:27

Acsor

Aug 29, 2019 15:26

I have an inquiry about a simple classical mechanics problem. Let's see if I can paste this picture

/dev/chat

General discussion for unix.stackexchange.com. If you have a q...

Acsor

Jun 10, 2018 17:12

Solved it! I just needed to install the `linux-headers-4.16.0-2-amd64` package, providing the kernel headers. I got there by looking up my kernel version number into the `aptitude` search.

Description-en: Header files for Linux 4.16.0-2-amd64
This package provides the architecture-specific kernel header files for
Linux kernel 4.16.0-2-amd64, generally used for building out-of-tree
kernel modules. These files are going to be installed into
/usr/src/linux-headers-4.16.0-2-amd64, and can be used for building

Acsor

Jun 10, 2018 16:31

If anything comes up that needs troubleshooting I'll probably head over the Q&A

Acsor

Jun 10, 2018 16:31

Thank you for the help and the cozy conversation. I was hoping to get some physical exercise today

Acsor

Jun 10, 2018 16:30

Yes, I meant: the stable is 9. Buster, as I may call mine, a preview of the tenth release

Acsor

Jun 10, 2018 16:28

Oh wait, we're on 9. So that makes sense. I'm left wondering of any codenames for the testing distribution. But probably it changes over time

Acsor

Jun 10, 2018 16:28

Well, probably the currently makes a difference

Acsor

Jun 10, 2018 16:27

It isn't clear whether that's a codename of an historic release or just to the testing distribution. Here's what I read on Wikipedia:

> The testing suite is currently named Buster,[125] which is the real (not the toy) dog seen in Toy Story 2 and Toy Story 3.
> Debian 10 freeze is planned for March 2019, and the release "some time mid 2019" and will be called Buster.

Acsor

Jun 10, 2018 16:25

@FaheemMitha Seconded

Acsor

Jun 10, 2018 16:24

Well, you see. The testing distribution must have been for a reason - arguably the dated packages of the stable distribution - and I'm not sure if I'm recollecting the right one