/sci/ - Science & Math

[Anonymous No.11317720]

>have a CS degree
>apply for interesting AI / data science / machine learning jobs
>every time the reply is "Sorry but we're looking for people with math/stats degree"

What the fuck is going on?

[Anonymous No.11317751]

>>11317720
Get a math degree brainlet

[Equivocation No.11317843]

>>11317720
Learn linear algebra you fucking brainlet

[Anonymous No.11317962]

OP you're a retard if you're really surprised. I did a stats degree with a CS minor and I always get offered jobs.

[Anonymous No.11317982]

CS fags BTFO

[Anonymous No.11318090]

>>11317720
Fakest thread I've ever seen in my life lmao. Every ai/ml/ds job is exclusively CS degrees. And none of them accept you because they all require either MSc (engineer levels) or PhD (anything else). Not a single ai/ml/ds job even so much as look at someone with a math or stats degree, it goes straight in the trash. The only non-cs degree considered is physics, and only for shit like applications to physics, or quantum ml, etc.

[Anonymous No.11318105]

>>11318090
>Not a single ai/ml/ds job even so much as look at someone with a math or stats degree,
not true at all. stop being such a butthurt CSfag

[Anonymous No.11318115]

>>11318105
Lol'd hard, thanks for the entertainment, clown-chan.

[Anonymous No.11318445]

>>11317720
I mean, do you have the math background? How much mathematics are you comfortable with? Everything that's interesting in CS is locked behind sophisticated mathematics. Also
>not double majoring to dunk on /sci/ anons
Having academic funding or high salary while still taking the same 3 semesters of topology, algebra, etc. as math majors pisses people here off to no end.

[Anonymous No.11318454]

>>11317720
>>11318090
>>11318105
Positions in ML/AI almost always call for CS with a math heavy background or for a software dev to help core implementation. They generally do not lock by degree - most of these positions have phone interviews where they'll ask you a basic ODE question and maybe 1-2 other easy questions to verify you know the basics. Then you have a few more interviews that test some of your actual math skills, alongside an interview for technical competency in writing algorithms, etc. After that, they decide if you're qualified.

Recall that many positions hiring for your level aren't even looking for the rigor, ESPECIALLY if they're data science jobs. Those positions call for a PhD, and at that point, having a PhD in CS (usually theory concentrated in computational learning theory,), Math, or Statistics is what they look for. This is mostly a thread to rile up
>hurr durr CS degree
with shitposting

[Anonymous No.11318455]

>>11317720
AI _IS_ math.

ADVANCED math.

[Anonymous No.11318456]

>>11317720
>>11318090
Can both of you post some proof of your claims? I don't believe either of you.

[Anonymous No.11318461]

I got a master's in economics and have ai data science/machine learning/deep learning position.

I did a couple of courses on Udemy and manage to take the job.

[Anonymous No.11318481]

>>11318455
Yes, as is TCS. As far as industry is concerned, they're not using the advanced mathematics to study ML. Most industry applications work around fitting models to situations and doing empirical studies. You need more analysis and algebra to study more interesting ML. Most ML/AI jobs aren't interested in that, or they're a lab.

[Anonymous No.11318488]

A CS major would try to build the AI and realize it's not feasible and doom the project.
A math major could spend decades working at it without ever producing results.

[Anonymous No.11318515]

>>11318488
These are both wrong. CS and math undergrads often study AI in grad. They end up studying the same mathematics and same computational theories. They end up doing the same research together. Background diversity is a good thing for PhD programs.

[Anonymous No.11318543]

>>11318456
I worked ml at m$ and every big name in ml try to scout me constantly despite 0 math background. Don't know what to give without doxxing myself but simultaneously showing some proof. You could google up some open job listings but we all know that never really means anything.
>>11318515
This is correct, but my point was specifically with regard that in either case you end up with a cs degree, not a math degree. There's no "math AI" degree, for example. In terms of undergrad backgrounds of people doing AI, it's anywhere from physics to math to cs to stats to economics, but I have never heard of someone with a non-CS higher education degree doing ML, everyone employed anywhere has <insert undergrad background> + either CS master's or <insert MSc background> + CS PhD.

[Anonymous No.11318547]

>>11318454
You have 0 clue what an ML position interview looks like and you have 0 idea how profiles are evaluated for these positions. Not a surprise though. This is /sci/ after all.

[Anonymous No.11318554]

the CSfags in this thread are fucking retarded as usual

[Anonymous No.11318562]

>>11318554
Lol butthurt neet.

[Anonymous No.11318572]

>>11317720
This is actually true, there's a lot of things that you need to be aware of.
This faggot >>11318090 is an actual butthurt CS nigger too dumb to do actual maths.

t. ds master with stats degree

[Anonymous No.11318574]

>>11318572
Nice picture of yourself, larper.

[Anonymous No.11318583]

>>11318574
shut the fuck up brainlet
go learn a new js framework or some shit

[Anonymous No.11318587]

>>11318583
Yum yum delicious larper tears. Cry moar, I hunger for a snack.

[Anonymous No.11318592]

>>11318587
I'd love to see your dumb ass train a CNN, cs dweeb

[Anonymous No.11318597]

>>11317843
if any cstard learned linear algebra thoroughly, developing true intuition they would cease to be a cstard and lose all interest in that subject. the same goes for probability and measure

[Anonymous No.11318600]

>>11318592
That's...
Perhaps the most embarrassing post I've ever seen.
You apparently think that typing "model.fit(X, y)" is something that requires...
What even?
Intelligence?
Skill?
I mean, it's really amazing. Do you actually unironically believe that "train[ing] a CNN" is what ml researchers do?
Ouch.
I mean, wow.
Even engineers technically don't do that.
Oof. I legitimately have no words for how disappointed I am with your post. I thought you'd try harder to continue your pathetic larp. Wow.

[Anonymous No.11318608]

>>11318597
Linalg is a mandatory prereq in every cs program though? Same for probability (and statistics).

[Anonymous No.11318978]

>>11317720
kek stat major larping
dream on and enjoy doing accountant tier work for the rest of your life topping at 70k/year

[Anonymous No.11319185]

>>11318547
>You have 0 clue what an ML position interview looks like
Lol, this was the process I went through. Care to elaborate your own anon?

[Anonymous No.11319431]

I'd your degree is anything less than Masters you are not qualified. Those fields are still not applied and are therefore closer to pure stats

[Anonymous No.11319464]

feels good to have done a double degree

[Anonymous No.11319675]

This was a bait thread. I played you guys like a damn fiddle.

[Anonymous No.11319696]

WHO DO I BELIEVE AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

[Anonymous No.11319735]

>>11319675
It means you can be ML engineer just with CS masters?

[Anonymous No.11319834]

>>11319185
Nevermind desu, I quickread your mediocre post. The only part that's retarded is muh ode since you couldn't be less relevant to ml/ds/ai. Prob, stats and linalg questions, on the other hand, are all very common, so that you brush these aside yet mention ode whatsoever is extremely odd and sounds like larp. But the rest is true except that they also look heavily into where you got your degree from (specifically, what lab, and if you don't come from either a reputed lab or a reputed school, you're fucked).

[Anonymous No.11319838]

>>11319735
Yes, ML engineer roles are usually master's (because they're some of the most boring and mind numbing positions ever, but at least they pay hand over fist, anyone with a PhD and half the competence to go with it goes for an ML research position instead).

[Anonymous No.11319839]

>>11317720
This is a massive COPE post by a mathfag.

[Anonymous No.11320052]

>>11317720
>did business degree
>join tech company as data analyst
>promoted to DS

What now, /sci/

[Anonymous No.11320585]

>>11317720
Really? I have a degree in Mathematics and "specialized" in stats and I also have self studied ML as well (have also complete Stanford's course on coursera to show). I actually understand the concepts very well I have no real experience on programming real world shit except basic procedural programming and some computational shit mainly on Matlab.
Should I apply to such jobs? You think they'll hire me?

[Anonymous No.11320752]

>>11320585
For engineering jobs no, since those are 80% writing scripts for data collection/cleaning. You should aim for research positions (and also learn Python well).

[Anonymous No.11320924]

>>11320585
Try it, you'll see the truth about OP's shitpost.