Is there some flag to close because this is essentially "do my homework" or "I am in Formal languages class [...] problem statement" ? Or should it get proper answer despite lacking effort whatsoever?
@EvilJS, I'm suggesting you add citations to your answer (not here in chat) so that if the links stop working, people can still find the papers you are referring to. As far as where it's published, have you searched for the paper title on Google Scholar? That will often tell you where it was published.
@EvilJS, maybe you've already added it to the answer? Let me see if I can dig up where that was again.
Ahh, I see, you added it to the answer. Cool, exactly what I was suggesting. Thank you, that looks very helpful.
@D.W. I know, this was first what I did, just wanted to confirm is it appropriate, and to check if summary is ok, it mayhap I should elaborate. Still learning how to answer because model here is very unique.
Question:
Is there established procedure or theory for generating code that efficiently applies a matrix-vector multiplication, when the matrix is dense and filled with only zeros and ones? Ideally, the optimized code would ideally make systematic use of previously computed information to reduce...
re the Trifonov ref: afaict the question is not actually over GF(2) as the paper studies. GF(2) has addition and multiplication as binary XOR/ AND resp, the operations are closed within binary values. the above question is matrix multiplication of binary matrix operands using std multiplication/ addition and/ but results are not nec binary. there is a lot of research on GF(2) matrix operations in cryptography but afaict, it does not apply here. — vzn1 min ago
@EvilJS oh yeah, ok. mobile is funky that way, have used it myself.
@vzn yes, checking if such ref is ok, and explicitly gave at (@), if something is enclosed over finite dictionary it is ok, also cannot use binary only operations as vectors are probably real.
@vzn I just did drone crash test. I knew it is called crash, but it literally crashed ;)
Lol. Friend was so proud, putting weights anywhere, self stabilizing systems, it was stable with pressured air from any side, self stabilizing in case two engines went down. And friend was so proud of deflect / dash system, tried throwing paper balls at different speeds, none did hit
So I took kniefe and thee it. Headshot, crash landing ;X
@EvilJS no kidding eh? have been tracking drone developments myself for many years. try this cool/ glossy new mag rotor drone. what drone are you using?
?!? throw knife at drone? are you crazy? sounds like an enrique iglesias move to me :(
@vzn no, no kidding, I threw a knife, it cut some wires. No real harm done, but still, did not passed my crash test. Yes rotor drone, it has no name and friend is doing prototype, so it is the only one manufactured.
@EvilJS ah yes. a lot of custom work in this area. but frames & motors & cpus are typically common. what kind of controller? frame? there is a lot of 3d printing going on in this area also.
@vzn I do not know this music, sorry. Well he said that dashing system is very mature, perfect, so I helped him with testing. Do not know details. I was only writting some quaternions and cheering him up ;>
It certainly make perfect dash when bird appears, but smaller object, not so much.
@Seanny123, tough to say, without seeing the question. I see two pitfalls here. First, if you're not careful, this could end up being too broad. We want questions to be narrow and focused. So, a question like "What is the state of the art in network flow algorithms?" would probably be too broad.
Second, it's important to do research first before asking, and show us in the question what research you've done. You should have done a literature search (e.g., Google Scholar). If you just ask for the state of the art with no prior research, there's a risk your question could have downvotes rained on it. See cs.stackexchange.com/help/how-to-ask.
That said, if you keep it within the rules of Stack Exchange (be clear what your question is asking, don't be over-broad, show your research), that kind of question could be entirely suitable and could make a great question for this site.
@Seanny123 hi/ welcome, curious, what are you doing at CNRG? have blogged on related topics, eg AI etc... (coincidentally banging on massive link pile for new deep learning post at this moment...)
@vzn yes, you are right, I am profusely sorry. Paper is about binary matrix and binary vector, I did find this scheme useful and treated it like finding decomposition idea, but did not provided that info, and haven't implemented and tested the idea yet. I am deleting it for now.
@EvilJS dont really think you should delete the answer. you could put a comment or an edit. think sometimes answers that are not correct but show a misconception are still worthwhile to keep. (it can point to unclear questions and help clarify questions iteratively. there are other cases of this, its not a new precedent.)
(sigh) see you deleted your comment, leaving mine referencing yours no longer making sense. :(
@vzn at this point I'm trying to work on using reinforcement learning so a production system can solve some simple puzzles. Basically, my Master's thesis end goal is to answer this question cs.stackexchange.com/questions/45651/…
