[ art / civ / cult / cyb / diy / drg / feels / layer / lit / λ / q / r / sci / sec / tech / w / zzz ] archive provided by lainchan.jp

lainchan archive - /sci/ - 72

File: 1476354001525.png (11.42 KB, 300x225, hqdefault.jpg)


So, what are your favourite areas of research?

While I like anything CS related, I definitely prefer more networking and security related research.


In particular I like statistical analysis of data sets so that goes into bioinformatics, or any big data projects for that matter. But in particular I like optimizing the fuerk out of the stack I employ.

My particular stack that has pretty interesting presentation capabilities is:

Kafka, Spark, Fabric, and many other setups make wonderful clusters that can do fantastic things. Mainly, croudsourcing difficult computational tasks are of interest to me. Historically, the breaking of RSA/DES and how they gathered a group of people and attempted to optimize the task.


Seems people eo not understand what the science board is suopose to be. Talking about "stacks" isn't it.


>The network stack

Outside of computers I'd say:
>Number theory

Maybe the last one sounds too political, and I'm not even sure it has a place in academia (I think it's called something like public relations, in the marketing world).
It's a shame that such an interesting field hasn't been properly studied because of the odd bell it rings


I like AI but barely understand it. I'm thinking about maybe studying it for realsies.

I like the most theoretical parts of physics, and I wish I could wrap my head around GR. The more applied aspects of physics(especially thermo) always have just a little too much handwavium in its proofs.

Crypto is fun but I'd never make a good cryptographer.

I have lots of friends who are super into language design and it's sort of rubbed off on me.

Besides that, I just love seeing a really elegant mathematical proof. Any sub-subject of that is fine with me.




>Maybe the last one sounds too political, and I'm not even sure it has a place in academia (I think it's called something like public relations, in the marketing world).

>It's a shame that such an interesting field hasn't been properly studied because of the odd bell it rings

Propaganda is definitely academic and has been studied extremely well. Though a reddit link it contains a massive amount of texts you SHOULD read, classics in the study of propaganda: https://redd.it/2gi04s


F.ucking A! thanks.
thanks again


File: 1476382126918.png (514.21 KB, 200x173, 1452715095419.png)

Programming Language Theory


Mostly functional analysis. I mean linear algebra meets topology? Sold.


I hated thermodynamics when we studied it at a semiclassical level. Coming at it from statistical mechanics/quantum mechanics really improves it. I actually enjoyed some of it.


Yeah I did the OCW course on information and entropy and I was very much more interested. Very good course.


Nothing. I have to force myself to study. If theres any good thing that comes from something so mindnumbing and boring, it's the end result.


Right now the science board needs activity..


>Functional Analysis
God-tier subject, I can't wait to learn it.

I'm mainly interested in algebra (especially Lie theory), analysis and mathematical physics (QFT looks rad).



>>73 Here.

Stacks is how I accomplished the science. You have to be some purist glitter boy to think that not looking at the intersection of analytical logistics with the pure science doesn't merit discussion from a point of view of hard science.

That in itself is a branch.


Learning processes, linguistics and others.


I like to study robotics, artificial intelligence and computational biology. I am building a laboratory dedicated to the subjects. I would like to do what HHMI did with the janelia institute and create a place dedicated to science having free reign to explore challenging and novel research without the drag of short term payoff focused grant reliance.

The research has been a part of my life for a long while though...I have been poring through articles about new advances since I was 8 or 9. I could barely understand anything lol.


Physics, particularly in the areas of plasma physics, turbulence, and QFT. More specifically, I like to read about quantum computation and nuclear fusion, but am still learning all the basics. Theoretical physics is wonderful too, but again I feel like I can't really appreciate it fully yet without having learned everything. I'm also diving into machine learning / AI recently, which has been really entertaining.

I feel you on wanting to break away from the publish-or-perish environment academia has seemingly evolved into. How would you even go about starting a lab like that?



I guess ill try to replicate success. No need to reinvent a perfectly good wheel.



some more about them and what they do. Seems like a pretty good model. They are still infantile but i suspect we will be seeing major developments out of their labs very soon.


That's neat. I was reminded of Bell Labs and the Institute for Advanced Study when I read your first post, so it's not surprising to read about the former being an inspiration for the janelia institute.

Regardless, that's a pretty noble endeavor, and admittedly I've thought about doing something similar. A source of funding always ends up being the issue though, and it's also part of the reason I have anxieties about going into pure academia.


academia feels like its more a place where good scientists go to die in my opinion.

But that would be discounting an awful lot of good work too.

You should definitely connect on the irc, Talking more about your ideas could prove helpful. I am on my way to building the funding. I have a plan and am actively working on it. May not be as big as HHMI, but I will get there. It is after all the only thing I want to do it life. lol
*nervous laughter*

I am Mahydraal, send me a PM sometime.


I sent you a PM yesterday via hexchat but i'm new to IRC so i have no idea if you got it tbh. I'm Bubbles!


I am very interested in Computer Vision and Machine Learning


oh! sorry, I must have missed you, my weeks are hella hectic


File: 1477840483339.png (78.45 KB, 200x93, amb.png)

>what are your favourite areas of research?
Anything related to compiler programming. Currently I'm studying parsing theory and automata and language theory. Due to this, I got interested in mathematical logic and theory of recursive function.

What is it about? Once I was reading an ancient book on that and it seemed to be what now is known as control theory.

>Functional Analysis
What I liked most about that is the elegant and simple proof of the existence and uniqueness theorem for ODE.


I don't know about control theory, I'm going to look it up.
But cybernetics is about the study of ... systems. Of the structure of systems and the roles of their components and how they interact.


I'm interested in neural networks, but getting anywhere near it will prove difficult for me. It shouldn't be too hard for you, however.

I found this course:


And this is apparently one of the lectures: http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf

Maybe this will be of some use to you, if only to see if you find the subject matter appealing. I can't handle it because my background in maths is shit. I have to teach myself calculus while working my pathetic code monkey/help desk job, hoping to one day touch AI. The only thing I have helping me is sheer obstinacy.


File: 1478226498674.png (1010.92 KB, 200x200, popek_virtualizable.pdf)

>cybernetics is about the study of ...
Have you ever study ancient papers on CS?
Do you know what a system is (in the formal mathematical sense)?
I've been studying if there is a mathematical model describing a virtual machine, and the most resembling information I found was this paper, in wich you can read how the authors use the term "system" but never define it.


Are you looking for abstract machines?



not too academical stuffs but research about practical attacks that we can implement right now.


I took a look at this course a few week ago, too. And you are right. It's quite theoretical at times. In the forums they recommended to first do Andrew Ng's Machine Learning course (https://www.coursera.org/learn/machine-learning).

I can only recommend this one. He takes a lot of time for explaining the calculations and gives advice for implementation and also a nice Octave/MATLAB tutorial.

I'm at week 4 at the moment where he gives a short introduction to Neural Networks. I think it is easier to follow Hinton's lecture on neural networks after this one. You will already know some basics.

As for me, I am more interested in the applications of (especially Convolutional) Neural Networks. The computer vision group at my university does some nice things with these. Therefore I will finish the machine learning course for the basics and then play around with some framework (Caffe sounds good) and read some papers.


Signal processing is a neat topic. I know it's super broad, but like everything in that domain is just so fucking cool.


Hell yeah!
Nothing wrong with enjoying a super broad field, I'd argue the opposite.
It's neat to understand general knowledge, as it leads to a large number of applications.
Just look at graph theory: nearly every practical problem can be expressed as one, speaking hyperbolically.

Not that fond of signal processing myself, but I feel that we would find common ground somewhere as I'm really into control theory at the moment.

I myself getting my feet wet with AI and generally how to implement it efficiently on a FPGA/ARM SoC for some robotics project at campus.