Hey Jow Forums

Hey Jow Forums,
I'm a first year CS/CPE student and just had an interview with some startup and they said to focus/spend my time learning better stuff.
I'm first year and wanted an internship this year so i figured i should self learn web dev shit, i actually just know html css and some django but whatever, and this interviewer guy said bro you need to do other shit, shit that you like, not just stuff that is easy to do.
So what can i start to self learn/practice that will impress interviewers (or whatever i should learn that isnt webdev), i have taken programming in python, intro algos and ds in C, digital systems, computer arquitecture and lin alg and calc 1-2, what do?

Attached: CS.jpg (890x670, 195K)

Other urls found in this thread:

youtube.com/watch?v=aircAruvnKk
coursera.org/learn/machine-learning
youtube.com/watch?v=2kupe2ZKK58
en.wikipedia.org/wiki/Pipeline_(Unix)
twitter.com/SFWRedditGifs

>Jow Forums g coding challenges
guess what you're supposed to do with this

>he fell for the programming meme
good money but it's soul-crushing.
you should have gone for management or working with other humans.
if you want real results to intentionally torture yourself start learning FORTRAN or ASM.
Or start seriously considering C and its derivatives.

Unironically read SICP

That's fucking terrible advice please don't try to advise anyone again.

OP, try to contribute to an open source project, something you would like.

Also don't seek career advice on Jow Forums, even Reddit is a better place for that.

>reddit
kys

>wants to get hired during first year
>html, css, django
You're on the right track for becoming a code monkey. Seriously kid, maybe try paying attention to your fucking courses instead of running after easy corporate money!

Attached: 1554133157398.jpg (450x399, 140K)

hey OP, fellow CS fag here. My advice is keep your academics up and try to take what you can out of university, but also find something that is niche and generally interesting to you, and publish your project to github so you can show employers your own passions

1) learn a language that is in common use
2) build a portfolio (you're going to need to do this regardless if you go to college)
3) get hired to work at marginally ok job
4) ???
5) profit

havent thought about that, from memory think its a bit hard but ill try to find the easier ones and see if im at that level.
currently learning ASM in computer arquitecure and C in intro algos and ds, also not really interested in management, not now at least.
i have heard of this book before, and if i can't find an internship for this summer, i was thinking of reading it, but again this kind of is on the same level as learnign web dev shits instead of something that will suprise employers, no?
>contribute to open source
yeah i was thinking of doing some kind of project, recently rewatched the first 3b1b video about neural networks and i was thinking if i could do something related to a program guessing which number its drawn on the screen, problem is i've only learned how to 'draw' or 'use a screen' in assembly, and its prob going to be real bad.
I sometimes browse cscareer, thats why i wanted to get and internship frist year just to have a massive headstart compared to my peers

Just wanted a summer internship, not real work.. Im planning on getting a masters, currently my grades are personally trash but im guessing im like top3 in class (out of 60), and uni is prestigious so you get job after finishing degree even with minimum grades.
Yo since you have '5g' in the name, would employers give a shit if i self learned some computer networks? Its 3rd year 1st semester class here, or 2y2s+3y1s+3y2s if you go for the telecom path, i also want to say again its not full time job, just a first year summer internship

depends entirely on what you want to do with your career. There is a lack of qualified programmers who know how to create an efficient netcode for vidya, so if you were willing to really learn those skills you could probably make a good living.

You're not going to get a good internship unless you actually know a little bit of something first though, as otherwise you're most likely to waste your time as well as theirs by accomplishing little more than nothing, unless you're able to soak up shit as you go

>depends entirely on what you want to do with your career
Im not really sure either, i like CS and maths, im planning on changing my degree to computational maths, taking all the CS related electives and then doing a masters PROBABLY on algorithms, im not very sure if this counts as what i want to do with my career because as of my understanding algos are required in every business so i can't 'plan'.
And you're right, i imagine a company would want someone who already knows a bit, which means hes interested and you both can gain a lot from that internship

javascript, mongo, node, react, graphql
don't like that? then go for ML with python

are FPGAs something that's available to learn where you are ? If yes, apply for that the first chance you get.
VHDL is like COBOL for young people : no matter how bad you are, if you can write it on your resume there's someone out there willing to pay you a ton of money for that

Attached: meanwhile.png (1500x1160, 1.32M)

I'm confused about what a ML project is/should look like, do you just say oh i manually gave 'it' commands and told it what the answer should be, or program a program with a shitload of if statements that feeds the ML program so it learns? Im not sure how can i dive deeper into this either
Just checked and i have a company that i really like/liked, but they offer jobs not internships, and they ask for 3+y experience, and verilog knowledge; also i kind of want something more related to software or maths because im planning to changing to computational mathematics

this kind of stuff is very interesting to me. I'm not op, but im wondering if i want to get into FPGA stuff should i learn specific stuff like EE, x86, C etc?

no titties bump

you probably didn't show enough enthusiasm/depth while talking about your project hence the "do more of the stuff you like"

current state of ML and """AI"""
you have LARGE dataset you want to classify
you have 2 basic ways of going about this, and their variations but forget those until you like it enough
1. Supervised learning - like you said, you need a bunch of preclassified data, you pass it in and it trains based on that. Then it will become decent at classifying data it hasn't been trained to classify via interpolation into the sets you trained it on. EG: classifying handwritten digits.
2. unsupervised learning - this will classify autonomously based on some criteria, but these classes will be unmarked, all the network will tell you is whether a piece of data is closest to some set. which afterward you yourself will label.
if you think you are interested in this you should check out some primer videos on ML, or even do the free course on coursera for it.

Hey, yes i think it's cool that a computer can 'learn', and im kinda mind blown by genetic(?) algorithms, the ones that try different things then select the best trait and keep to changing other things, and i also like the general idea of optimization and efficiency, which i dont really think is related to ML but is something that i think about.
>if you think you are interested in this you should check out some primer videos on ML, or even do the free course on coursera for it.
i dont get what's primer, i googled and it seems to be a introductory textbook, is that what you meant? And yeah im going to check coursera, at least the free courses because im the typical poor uni guy; do you think i should know/learn anything before reading a ML textbook or can i jump straight into it? Thanks btw

should have elaborated, by primer i meant like really basic youtube videos just to get your head around the basic concepts, after that the coursera course will take you a bit deeper and explain how programs actually work underneath (a bunch of matrix multiplication) and why they work like that. The coursera course has been free for like 10 years old so no need to worry about money, and it's pretty comprehensive, if you can't immediately understand all of the math, that's okay it gets better with time (and isn't REALLY necessary to build models). I'm not aware of any good textbooks though, so i can't help you there, sorry.

by primer i mean stuff like this: youtube.com/watch?v=aircAruvnKk
the course i'm talking about is coursera.org/learn/machine-learning it's basically coursera's flagship courses cause I think it's done by one of coursera founders (don't quote me on that it's been too long)
also remember that the ANN is only the classifier part, which can then be combined with other algorithms for greater effect.

also by genetic algos do you mean those ones that do a bunch of simulations and then pick the most fit ones to use as the seed for future iterations? if so here's an example of both an artificial neural network (albeit EXTREMELY simple) and evolution algorithms
youtube.com/watch?v=2kupe2ZKK58
there was another better one but i can't find it, evolution algorithms are pretty cool though

>by primer i mean stuff like this: youtube.com/watch?v=aircAruvnKk
Yeah i watched that video like 5h ago, its what made me make this thread other than the interviewer making me feel like im wasting my time.
I already signed for this coursera course, im just worried that it will be too slow, and wanted to start looking for a book to check during the summer but ill try to found out, also i dont use my unis book because everyone says our AI class is trash.
And yeah i meant genetic algorithms, made me mind blown when i first saw a video about it, like 2 days ago i rewatched a 3min video of something learning how to jump over a ball over numerous tries.
Thanks for answering man, can you add me on discord so i can message you once i have a clearer path of what i will do? frazzles#4215

why do you want to do a masters? If it's for going into academia, you are wasting your time with web dev.
If it's for career advancement, then if you want a job in webdeve then fine.
If it's for "prestige", you are wasting your time and money. Remember that most universities are now sudo scams: they teach you stuff under the guise of their old reputation as centres of learning and discoveries, but the majority are businesses to the core, especially at the masters level (less so PhD or undergrad's, funnily enough)

Im finishing first year of CS and planning to changing to mathematics next year, im planning to do the mathematics algorithms masters, because its just 2 more years and thats usually when companies offer relatively huge salaries around here.
I did webdev in the hope of getting a summer internship during my first year and then being easier to get another summer internship in the following years, im not particularly interested in webdev.
And i think that for a more advanced field you need a masters, unless youre autistic, and im not that autistic.

lol i'm not an expert not by a longshot, i just know the basics, so i don't think i can be much more help. but what i can say though is most current ML jobs have more to do with building models using already existing libraries, so along with doing the theoretical stuff i'd also suggest making a simple model with something like tensorflow in the language of your choice so you can claim familiarity with it and turn heads on interviews. Also since you're only in first year this is likely already leaps and bounds above your competition (although it depends on situation i guess) so just being passionate in talking about it and knowing many details should help a lot.

you will learn a lot more from interacting with actual ML related devs, much more than you would learn from a book or from an imageboard, and even from doing projects by yourself, both in terms of experience with ML development and in terms of working in an industry, interacting with other engineers and business personnel. the only other thing I can say is try to have multiple small internships while in school instead of one big one, so it can give you more exposure to different teams and styles, and even different fields. and if it can be helped at all, NEVER work for free.

Good luck! Fight hard and you'll make it!

>problem is i've only learned how to 'draw' or 'use a screen' in assembly, and its prob going to be real bad.

It's way, way, way, way, way easy.

t. rendering engineer.

Thanks fren, i have a first year friend that does some kind of internship every tuesday about machine learning, but hes way too autistic tier, cant even compete with that but ill try to join there next year, as its part of shcool.
I've only learned how to write programs, not exacly how to access peripherals (you need to write on the pixels on the screen right), and im also interested on this, like a year ago i saw a guy post on Jow Forums that he set his caps lock LED to turn on when the computers temperature is above x degrees

>set his caps lock LED to turn on when the computers temperature is above x degrees
that is way easier to do than you think, provided youre on linux

I have windows and ubuntu installed on my laptop, ubunto was needed for C class, but i dont really know anything about the terminal, just cd ls rm mkdir i think, where can i check more about this? Did he get the memory place where the caps lock LED is stored and set it to one if temperature is high? But how did he check temperature?

you learn by using, stop using windows start using ubuntu and eventually you will run into things you want to do but dont know how so you start researching online
if youre in college you should be learning how to figure these things out yourself and thats important since thats 95% of any engineering job, software or otherwise

depends in the country you are in frankly. In the us you will be fine without a masters if you're good. in the EU you might be career-blocked in certain countries (notably France and Germany).

Im too far from becoming a Jow Forums hackerman, what i learned is mostly from reading how to install stuff and with minimal attention, id rather learn it from a structured approach like a book but is it even a valuable skill to know how to use the terminal?

Portugal, and you can get jobs if you come from top unis, not exactly sure if you need masters but my professors keep saying you need a masters

yes, most jobs will have you using a linux box somewhere, whether on your desktop at work or remoting into a server etc
and no this isnt really something you learn in a structured way, its like riding a bike or programming, best learned by doing yourself

there is nothing to learn.

1. You need to know that Filesystem Hierarchy Standard exists
2. You need to know pipes exist en.wikipedia.org/wiki/Pipeline_(Unix)
3. You need to know they use a bunch of dumb fucking names for programs and options.
4. You need to know that GREP is a thing.

Congratulations you just mastered the terminal. In the real world someone from your job will give you an unreadable command string that just werks and you will copy and paste it because you have better things to do.

As i said im trying to go for a more mathematical related career, but yeah ill probably use it on the way, and i can flex on the bio girls with that
Typical Jow Forums chad, im just learning about pipelines in my comp arquitecture class, im going to be presenting some small googling about speculative reasoning that kinda exploits this pipeline (for a shitty diferent class)

if youre doing a maths career without a doubt you are going to be remoting into beefy linux servers to compute things, it would do you well to learn

Oh ok honesly had no idea, also can you point me to the right direction to doing something like the caps lock led turning on at a set temperature?

theres multiple ways of doing this, pick a scripting language and google how to get cpu temps using a library someone else built and then figure out how to interface with a keyboard using that scripting language
thats just one way

Ok another question, how common is it to use libraries? Isnt it a bad principle or something? I know this is related to reinventing the wheel but its strange.
Bonus:what makes a language a scripting language? Is it just because its easier to make scripts with it? And a script is just a small program that does something right

very common, if youre doing maths youre going to be using python with numpy/pandas a lot for example
scripting languages are usually interpreted instead of compiled but theres no hard and fast rule for them

Alright thanks a lot man im going to sleep now

Anyone?

FPGAs are programmed with VHDL/Verilog, you dont even really *need* an fpga to start learning as theres IDE's with emulators but you can do some research and find a cheap FPGA to actually load code onto and do things with

in uni as a comp eng my FPGA class was learning VHDL on a real board and we did things like hook up keyboard input, LEDs, and eventually got it to play pong with output to a monitor, so i reccomend doing something similar to that and googling around to get the answers
the IDE/toolset we used was from Xilinx but i cant remember which product specifically

though now i do absolutely nothing related to FPGAs so i cant really help beyond that

forgot to mention theres quite a bit of low level knowledge required, which you can pick up by learning vhdl/verilog but its VERY common to be doing low level bit manipulations, masks, etc so if you're not comfortable with that stuff get a book on it

Your best bet is to start some projects. Find something you want to make and make it. Contribute to projects you like. A good github profile goes a long way, apparently. The last company I interviewed with wouldn't shut up about how important this shit is.