1:30 AM
1:45 AM
2:31 AM
@vzn yup, you have it online. I did today 1n+1, 2n+1, gap, and then switched to 9n+1, as it seems to be first interesting to give some insight, also still thinking that base 5 or 11 would give me answer to my own conjecture (it is a bit weird to make conjecture about two conjectures to take a look at "the conjecture".
14 hours later…
5:07 PM
@EvilJS its "bursty". it has pros and cons. the pro is that you dont really have to worry about what you write in chat rooms. the con is that it hasnt really "gelled" like some other SE chat rooms. but on other hand not many SE chat rooms have "gelled", seems like only half a dozen or so at times.
5:40 PM
@vzn started nowadays to tackle again the problem. Not suprisingly I use JS. The fastest way, and the most of the time I have access only to mobile phone when I write (computer is saturated, lab computer also), so I write like in natural language anywhere I want.
These gives the fastest deployment cycle, also I do not nag about speed, it is blazingly good. Yesterday I started binary prime recognition, it is giving approx patterns, so I thought about making Collatz only on consecutive primes (starting points), and about base 5 and base 11, only thinked about it, not implemented.
I thought about one thing: if it is possible to show inverse Collatz, it automatically proves no cycles or divergent case? I would really love to involve math to the problem, but there are several topics I am good at, with practice, but formal proofs or advanced math theories - either I do not know them or I would mess something up. For example once I did get stuck on physics and mathematics, so I bought two books: physics for game programmers and mathematics for chemists.
5:53 PM
I am doing EEG on humans. Read their state, emotions and so on. But EEG is on the surface of head, better idea is to get electrodes strictly on brain surface. This is possible on mice or rather rarely when human already have need to do some transhead insertion. These cases are patologic, so I do not bother to find them.
I have been there, but there are cultural differences, language barrier. And my experiments are strongly advanced on methodology and own code. So it would be hard to find someone that does not belive in commercial software, writes on his own and will take patience in language bartier
There is for example one item test on self-esteem. It is working in U.S. and does not work in other countries. And also if someone is proficient in English and made diploma abroad it is not considered valid in U.S. One have to take university courses
EEG and Eye Tracking is rather independent on cultural differences, but again. Where are you from? Oh... geee... Ok... :-/
I will have chance to speak to them when I get more titles or my patent will be resolved successfully. My equipment is very good (but only as good I can afford, so I work as a programmer to cover research expenses). For many people these would be toys... So I have to improve algorithms and methodology, to squeeze as much as I can, because better equipment is not an option.
6:22 PM
6:37 PM
I had contacted with Paul Ekman (guru in emotion reading, like Donald Knuth in psychology) when he gave testing of software that he is currently involved in. But very politely gave us response that he gives it to only top companies in the field. Another attempts were to EEG companies to purchase equipement. They just did not cared about customers from abroad.
Even though we wanted really big package - 100 devices + some materials backups (like electrodes), and politely informed that it will be used on international music competition that takes place here. Not talking about discounts or something, straight deal plus free commercial if they would like it. Did not bother to respond to purchase, emails and so on. Sorry for lengthy response, but this is how reallity looks like.
6:51 PM
Emotions reading, and other mental states. No, it does not apply to therapy. I am trained in cognitive-behavioral therapy, and I am not connecting it with EEG.
During any event that we present technology - wy strictly don't give a .* about slides or convincing people, we just give opportunity to try it - and since it works, we do not need any training in advance, any particular subjects and so on.
Total freestyle, give stimulus you want, give questions you like to answer (if it is possible), we make proper stimulus presentation put EEG on someones head and after experiment we give results. We did commercials tests, hidden affect tracing and other commercial application, even UX testing, which shows false results when using Eye Tracker only.
And there were times that we got results that were strictly too much. We were asked to destroy evidence, as some people wanted to bet it is not working. They were so sure, that asked if it can be lying detector. Ok, no problem. But results not only showed what was going on, after showing results people were embarrased and admitted results to be true.
Since this is psychological evaluation of some kind, and we are bound on confidentiallity and other clauses, we cannot admmit that somebody was tested if he/she says no. We are obliged to destroy evidence if somebody says so. So we did. We do not need any reason, if somebody says it is finish, destroy data - it never happened.
So now we are not so eager to show any sexually connected stimulus and other controversial themes - we end up with data loss very often.
7:23 PM
6
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...
Also doing private research (of course with consent) doing therapy is out of option. It may interfere, so I never even considered it.
Even slightiest unprobable and unproven anyhow chance that something might have minimal negative outcome - never betting on that.
Once I had test on workers in company, everything was great, andI asked personel not to enter and did everything that was needed to provide privacy. And out of the blue boss of this company came, and loudly asked what they think about him. It was to late to stop him so I asked him to leave, checked results and it was clear that many of them liked him, but not everyone.
1 hour later…
9:21 PM
10:15 PM
10:54 PM
11:30 PM
If I want vacation I change wallpaper, still working solution ;). I do many tasks on phone, and since I installed several compilers, it's hard to get away ;). Imagine that: you go diving, than sunbathing, it's getting boring, so you write program, personal project, so it is still valid vacation then ;)
« first day (1271 days earlier) ← previous day next day → last day (3170 days later) »