/sci/ - Science & Math » Thread #11567720

171KiB, 720x350, 20200412_151137.jpg

View Same Google iqdb SauceNAO

Learning Linear Algebra rigorously?

Anonymous Wed 15 Apr 17:06:53 2020 No.11567720 View Reply Original Report

Quoted By: >>11567740 >>11567817 >>11568534 >>11568554 >>11569486 >>11570831

I'm a CS major with an obsession in learning stuff in a rigorous way. I'm currently choosing between Halmos' Finite Dimentional Vector Spaces and Hoffman and Kunze's Linear Algebra. What's the best choice out there?

Anonymous

Anonymous Wed 15 Apr 2020 17:13:55 No.11567740 Report

Quoted By:

>>11567720
Hoffman and Kunze. Halmos is the OG but the latter is better for self learning.

Anonymous

Anonymous Wed 15 Apr 2020 17:19:24 No.11567765 Report

Quoted By: >>11568548 >>11569486 >>11571470

Definitely Hoffman-Kunze — it's the standard for a pure math linear algebra course and after it you'll be able to cheese a course in abstract algebra if you ever want to take one. Halmos' book is more of historical interest and the writing style and terminology are kind of dated.
As a CS major you might be more interested in applications and numerical methods; if so you could complement H-K with something like Carl Meyer's "Matrix analysis and applied linear algebra".

Anonymous

Anonymous Wed 15 Apr 2020 17:31:59 No.11567803 Report

Quoted By: >>11569486

Artin, "Algebra"

Anonymous

Anonymous Wed 15 Apr 2020 17:35:18 No.11567817 Report

Quoted By:

>>11567720
H and K gives you a better grounding. Halmos is for the curious, and for first timers kunze is better

Anonymous

Anonymous Wed 15 Apr 2020 20:47:53 No.11568534 Report

Quoted By:

>>11567720
H and K is good. Artin is good to diversify and learn more abstract algebra - and there's a lot of there that is right at home in CS. Aluffi is also a fun read for more abstract algebra if you're interested

Anonymous

Anonymous Wed 15 Apr 2020 20:51:35 No.11568548 Report

Quoted By:

>>11567765
>you'll be able to cheese a course in abstract algebra
>cheese a course

What did he mean by this?

Anonymous

Anonymous Wed 15 Apr 2020 20:52:48 No.11568554 Report

Quoted By:

>>11567720
Where is this text from?

Anonymous

Anonymous Thu 16 Apr 2020 01:55:10 No.11569486 Report

Quoted By: >>11569610

>>11567720
>I'm a CS major
I am so sorry.
>with an obsession in learning stuff in a rigorous way
So you're going to abandon CS next year (or the one after that if it's cancelled) and become a math major, right
>I'm currently choosing between Halmos' Finite Dimensional Vector Spaces and Hoffman and Kunze's Linear Algebra
Both are good. Personally I like Shilov but all 3 are well recognized.
>>11567765
>As a CS major you might be more interested in applications and numerical methods; if so you could complement H-K with something like Carl Meyer's "Matrix analysis and applied linear algebra".
This but I would say that to any STEM major taking theoretical linear algebra as well.
>>11567803
He's not ready, don't scare him away.

Anonymous

Anonymous Thu 16 Apr 2020 02:46:43 No.11569610 Report

Quoted By: >>11569715 >>11569878

>>11569486
Your undergrad is showing.
Algebra by artin isn’t hard lmao. There are a lot of CS majors who ended up taking courses with it at my school. You should fear algebraic number theory, not babby’s first algebra book

Anonymous

Anonymous Thu 16 Apr 2020 03:17:30 No.11569715 Report

Quoted By: >>11569718

>>11569610
>number theory
I'd rather not waste my life

Anonymous

Anonymous Thu 16 Apr 2020 03:19:30 No.11569718 Report

Quoted By: >>11570368

>>11569715
Yeah cryptology is pretty useless

Anonymous

Anonymous Thu 16 Apr 2020 04:33:23 No.11569878 Report

Quoted By:

>>11569610
>There are a lot of CS majors who ended up taking courses with it at my school
Your European is showing

Anonymous

Anonymous Thu 16 Apr 2020 08:28:44 No.11570368 Report

Quoted By: >>11570500

>>11569718
this but unironically. every major data breach in the world happens because some dumbass thought he understood crypto but fucked up the implementation. either become a dedicated expert or leave it to the spergs and find something more useful to study on the side.

Anonymous

Anonymous Thu 16 Apr 2020 09:51:21 No.11570500 Report

Quoted By: >>11570792

>>11570368
That's completely wrong. Most data breaches occur because of ignorance towards cryptology, out of date software, and default configurations. Stop trying to pretend like you know what you're talking about.

Anonymous

Anonymous Thu 16 Apr 2020 12:33:45 No.11570792 Report

Quoted By: >>11570808

>>11570500
>ignorance towards cryptology
>out of date software
>default configurations
exactly the type of mistakes made by incompetents who should by all accounts have left it to the experts. especially the former.

Anonymous

Anonymous Thu 16 Apr 2020 12:39:19 No.11570808 Report

Quoted By: >>11570911

>>11570792
> cryptology = software security or network security
> "major data breachs are caused by wannabe-cryptograhers" but describes a problem related to implementation, usually done by clueless devs
> "should left it to experts" but has no clue who could actually be such experts
Cool story bro.

Anonymous

Anonymous Thu 16 Apr 2020 12:48:55 No.11570831 Report

Quoted By:

>>11567720
What it that from and why do they have such a confused understanding of philosophical concepts?

Anonymous

Anonymous Thu 16 Apr 2020 13:24:51 No.11570911 Report

Quoted By:

>>11570808
you write as if you know what you're talking about, so let me just remind you of all the clients that came to you with a dodgy-ass custom implementation of diffie-hellman, or their own proprietary key exchange or authentication scheme. or they use an established encryption scheme incorrectly and leak information right out of their cyphertext. you know what would have prevented all of this? hiring someone who dedicated their career to not making these sorts of mistakes and careless indiscretions.

look, I trust that you know a thing or two about the subject, but the dunning krueger effect is a dangerous thing, and it manifests exceptionally strongly in undergrads who take a course or two on a proprietary subject. years later they get a job in industry and you know what happens.

Anonymous

Anonymous Thu 16 Apr 2020 17:19:22 No.11571470 Report

Quoted By: >>11571878

>>11567765
Thanks I downloaded a pdf of it and it was pretty good. Would I be able to btfo memechine learning fags who only took basic calculus and strang's linear algebra with that background?

Anonymous

Anonymous Thu 16 Apr 2020 19:27:01 No.11571878 Report

Quoted By:

>>11571470
Memechine learning fag here. It'd probably depend. In my graduate courses, 70% of people took cs as their undergrad, and you'd probably be able to btfo them. The rest, math majors, maybe not. Machine learning is honestly a pretty independent field. Most of the stuff you learn about vector spaces and multilinear maps, not very useful. You won't encounter it in most machine learning papers. In ML linear algebra is just a framework for datarepresentation. Just a ton of matrix and tensor productions, whatever. Hardest math, probably, optimizations, is analysis stuff, tensor calculus. The abstractions you earn in a rigorous formulation of abstract algebra isn't useful when you're always working with the same data-form. Regardless, the answer is yes.

Capcode	All Only User Posts Only Moderator Posts Only Admin Posts Only Developer Posts
Show Posts	All Only With Images Only Without Images
Deleted Posts	All Only Deleted Posts Only Non-Deleted Posts
Ghost Posts	All Only Ghost Posts Only Non-Ghost Posts
Post Type	All Only Sticky Threads Only Opening Posts Only Reply Posts
Results	All Grouped By Threads
Order	Latest Posts First Oldest Posts First

Your latest searches