I am an university student very much interested in ml/dl .
Suppose i am doing some good projects and kaggle competitions. And after finishing some courses i want to apply for a full time job in a startup or any tech giants.
Will tech giants accept me as a fresher?
How important data structures and algorithms are for me? I know basics as i am from CS background. Do i have to go deep in it?
Do comapnies require a specific framework like tenserflow or i can work on fastai (i just love it) ?
Truly great companies (size rarely matter, and often the smaller the better) where contributors’ potential can be discovered and grow organically, look for not what you already know, but your ability to learn new things and solve problems you didn’t encounter before, and learn whatever necessary new skills or tools along the way, from whatever resources that you endeavor to seek.
Curiosity to ask questions, explore answers, validate/invalidate claims (yours and others), and ability to learn that count more than a degree or certificate.
Data structure and algo are just constructs to help manipulate mathematical problems. Some problems are truly mathematical in nature, but many tougher problems rely more on common sense, and common sense often lead to new data structure and algos not available decades ago.
Consider keep working on sharpening your senses, be critical to any claims and assumptions you hear. E.g. can all life problems be expressed/quantified by mathematics? are you sure? Do “smart” phone make us smarter? Is “new” really better? Will DL/ML make users lazier?
Accumulate experience by practice: solve problems that you find interesting and better yet, those are obstacles for others, and would be helpful to them if solved (e.g. would make them more productive, help them learn and see with clarity) . Solve a few of these with whatever solution you come up with, I doubt you’d have problem finding a company who will want to use your problem solving skill.
What will that achieve for you, being accepted by a big company? Pride in front of friends?
Are you sure you will get what you want from only tech “giants”, and not from other smaller communities? This forum isn’t run by “tech giants”, and many here probably aren’t necessarily working for big giants either, why do you come here? (you’re very welcome here, by the way, most folks here are nice )
I wouldn’t jump too too high there, but yes, some company will notice if you put in the effort. I basically started there, only taking the courses in the beginning and doing my own projects (not even Kaggle really), and I’m currently in the middle of my first internship (how I’ve seen you get into companies mostly in this field).
The more you know how they work, the more efficient your code is. IE, I want to run some feature engineering. What’s better? A giant list and a for loop? Or is there some better way of doing this efficiently. Without this knowledge, suddenly a 30-45min script could take days
Yes and no. You’d want to look for PyTorch if they don’t ask for fastai (as this is quite rare/up and coming specifically asking for fastai), or see how flexible they are if you want to stay comfortable. What’s important is you understand what the framework is doing and how it’s doing it with a decent understanding. This way if you’re in a situation where you can’t use the fastai library, you can still generally get around.
Take the fastai courses
Build (many) projects along the way
Apply for internships and jobs at an intro level Data Scientist (names get odd in our field, so don’t just do DS specific, could also be Software Engineer, etc depending on the company, really read into what they want in that person)
Being tier 1 doesn’t make you any more special. How about kids from other tiers, do they deserve less priority? Middle class compare to what? Throw that class idea out of your mind, and you will be able to access and talk to a bigger pool of mentors from all walks of life.
If you want to know something, just do it, don’t count on a forum to give you predictive answers, or hope others will spoon-feed you tips and shortcuts that guarantee some acceptance. Life is best explored (for your own benefit) when you find answers to your questions by doing, not by asking for promises. Straight forward enough?
I gave you the answers already in the first post. Remember that sometimes others questions are pointers to make you think harder then the answers will occur to you — may not be right away, and you have a whole life ahead of you to find out your own version anyway.
I never for a second thought your questions were stupid. But as I said, socratic discussion — answer by questions — isn’t for everyone. But if you get it, it’ll hopefully help you learn much faster than the rest.
I’m glad you clarify, that clears up a lot, thank you.
I’m sorry to hear that the social circumstances you are facing seems unfavorable if not outright unfair.
Consider this for a moment — if a company or a recruiter only looks @ labels such as someone’s degree instead of real ability and contribution, how wise are they? What does that say about their judgement and quality of work? Or, if their clients only look at tier-#, do you want to work with/for those firms or mindsets?
I once chatted with a traveling pension fund manager from wall street on new recruits he came across over past two decades:
“Those Ivy League kids ----they’re more or less all the same. Can’t think out of the box. Boring.”
Be original, and bold in tinkering problems (practical ones), even if you don’t know how yet — the whole purpose of the exercise is to cultivate an attitude of "exploring what you don’t know yet."
And don’t give up lightly. Jeremy once said (in a concluding video of course-V3, DL1 I think), “There are thousands of unfinished projects on github — don’t be that person. Start something, and finish it.”
The truly wise and sharp minds look for that kind of attitude, not “tier” labels — I’m repeating myself of my first reply. With some silly but original projects of your own, that speak louder than any “tier-1” label.
Rest assured there is no stupid question. No question, nothing learned.
Have I confused you enough? I hope not. PM me directly if u want to hear more, I bet others are bored to death by my long speech.
Hey! I feel terrible for the words i have used before. We both really misunderstood each other. Can i have your any contact detail so that we i you can guide me towards my goal?
It can be your linkdin account, email or whattsapp.