CS books thread
Post them
Springer is pretty based
This isn't a CS book any more than a calculus book is a physics book.
>>7841418
http://4chan-science.wikia.com/wiki/Computer_Science_and_Engineering
>>7841418
a classic
This is the parsing theory book.
>>7841488
I think he means textbooks that would be good for a CS major to read, not ones that talk about an area that is exclusively CS, so graph theory is fair game due to the huge overlap between CS and Math in it.
>>7841733
>huuurrr book has numbers, formulas, equations
>it's a math book
Nice reasoning faggot.
>>7841870
You know that both graph theory and mathematical statistcs existed before computers, right? I've read both of those books from cover to cover. Have you, faggot?
>>7841961
Also, the former is literally preceded by the header "Graduate Texts in [math]Mathematics[/math]". He's just mentally handicapped.
>>7841961
I only read the latter
>huurr mathematical statistcs existed before computers, right?
''computers'' existed before digital computers fag.
>>7841418
beginning dump
beware the books im about to post are all kind of plumber-tier, but its knowledge you need to have
otherwise its just theoretical academic circlejerkery
>>7842112
you need to know the basics of a linux distribution, no exceptions
>>7841727
this has been recommended to me so many times in my graduate work in eee. can u elaborate on what makes it so based?
>>7841733
What CS major thinks they invented graph theory? It's merely useful for CS.
Besides, the book covers things like computational complexity and many algorithms that are due to CS. Even the preface says there is a huge overlap and that CS has made significant contributions to the field.
>>7841418
>>7841733
>>7841961
>>7842489
what is this stupid bickering. have any of you really studied graph theory or encountered a real-life graph problem?
well let me tell you, it sucks, bigtime.
do you guys even know how weak the modern theory of graph truly is?
to be able to say vitually anything of interest about a graph you first have to compute some autismal, contrived property like the diameter, genus or set of primitives (if it even has one) which is virtually impossible to compute even for ridiculously tiny graphs.
even if the graph has a lot of intrinsic structure, like regularity or bipartiteness, youre basically SOL unless you know these retarded, contrived values. otherwise pretty much all you can say
> its um... er,, uh... its a graph...
EVeryone in the world who wants to study a large graph wants to find communities/clusters in it for dimensionality reduction......... except there is literally a paper that says you CANNOT formulate a precise idea of this https://www.cs.cornell.edu/home/kleinber/nips15.pdf
one of the MANY other MAJOR FAILURES of the theory is that there is virtually NO CONCEPT of 'graph nearness.' If you know a lot aboutone graph, then swap around only a few vertices or edges, you are instantly CRAPPED and know nothing about the new graph. its a completely different animal. and an ugly one too, because it is a GRAPH
even spectral clustering techniques (note the word TECHNIQUE not 'theorem' or anything that really says anything) work simply on the dubious basis that 'hurr, durr, matrices with similar rows are usually close to singular'
the theory of graphs is LITERALLY TOO WEAK to handle ANYTHING people in the year 2016 are trying to apply it to. it has utterly FAILED to accommodate its most important use case. there are simply too many pathologies for the idea of 'graph' as it currently exists today to have any meaningful contribution to humanity
graph theory is for the gays
pic related
>>7842532
>what is combinatorial optimization
>>7842545
ever wonder hwy all the problems in the field of combinatoral optimization have trite little names like the 'travelling salesman probelm' or 'minimum spanning tree finding algorithm'?
because NOONE HAS PROVIDED A DEFINITIVE, GENERIC SOLUTION TO THESE PROBLEMS using the graph theoretic tools people have worked so hard attacking them with
instead of actually using any of these useless graph tools, in practice people end up using non-graph-theoretic approaches such as simulated annealing
>>7842553
There's computationally efficient algorithms (and these are old af) for minimum spanning trees
Good optimization algorithms for TSP exist
>>7842553
>NOONE HAS PROVIDED A DEFINITIVE, GENERIC SOLUTION TO THESE PROBLEMS
Linear Programming
Integer Programming
Convex Programming
Mathematical Programming
>>7842674
youre proving my point mang
all of that stuff was invented to avoid graph theory
prove me wrong protip you cant
>>7842553
Can't find an actual use for bipartite graph modelling... Brainlet detected
>>7841540
Today Ive seen there's gonna be a free 1 week conference/workshop in Bonn
https://www.mpim-bonn.mpg.de/node/6459/abstracts
From the abstract of Vladi, seems he's trying to build a model of HoTT in ZFC to make it more attractive for the mainstream people. He's been complaining about too few mathematicans (as opposed to type people) being involved in that for a while.
>>7843286
That sounds awesome.
I totally feel you on the lack of math students in HoTT. All of the pure math stuff I'm actually interested in seems to be in the comp sci department. Meanwhile most of the pure math studied in the math department seems to just be applied math in disguise (the courses and material covered are very much determined by applications to science or finance which I care nothing about).
>>7843286
> build a model of HoTT in ZFC
Do you mean a model in the proof-theoretic sense? Aren't there already models (in this sense) of intuitionistic type theory? I guess the only hurdle is figuring out how to interpret univalence.
>>7843216
bipartite graphs show up everywhere
but the POINT of modeling is so that you can say things about the model and perhaps INFER or PREDICT things from it
and the theory of GRAPH is simply TOO WEAK to be able to do so!![math]![/math]!