ML engineer career advice

Question

Thinking about starting my career in Machine Learning (MLE)

So, I'm currently a 19y/o programming student, very proficient in C, learning OOP more seriously in the next months. Programming isn't a problem for me (I've been programming in python since the age of 13).

I'm very passionate about machine learning, especially deep learning / meta learning but obviously when I try to get my hands onto any aspects of this world a little deeper, I understand that's close to impossible to be able to even understand a paper without a solid background in statistics / applied maths.

Soo.. the big question comes. With the final goal of becoming a deep learning engineer (not data analyst), and therefore take a master's degree in AI (or "Data science for ML" [my university has both]), is it more advisable to start a bachelor's in CS or Statistics?

I'm intrigued by statistics and don't mind the math as long as it's going to be useful, and as already stated I'm not likely going to face problems with the programming side later as I'm already proficient.

My concern is that the curse in statistics has many exams (in the third year) related to economics and sociology

MLE3760 · Accepted Answer

First of all, although it might not have been intentional, you come off quite smug. "Programming isn't a problem for me" and "I'm already proficient in programming" is not the correct mindset to have, ESPECIALLY at your age. You "learning OOP more seriously soon" indicates that you're still near the beginning of your programming journey, not at the end! Are you able to consistently solve leetcode medium/hard problems in 15-30min? Are you an absolute expert in system design? For better or worse, these two skills are what get you the job in Big Tech nowadays.

And in general, there is no limit to this knowledge, there's always new features, libraries, languages and also just plain programming concepts coming out. My sincere advice here would be to always stay humble and hungry for new knowledge!

Now returning to your OG question, CS sounds like a better path for your bachelor. And in general, if you're really interested in ML, I would strongly encourage you to just learn about it yourself in your free time.

Pick an ML concept and research it until you understand it so clearly that you could explain it to someone who doesn't have any idea about ML. Then try to implement it from the ground up (without using many/any libraries) and get it to work. Then try to slightly modify it. Repeat for a new concept. If you keep doing this, and at some point, are able to implement SOTA papers using just C and CUDA, I guarantee you will be lauded as a tech genius/prodigy and companies will literally be fighting for you.

adk97 · Answer

Hello, AI engineer here. Id go with CS because majority of your workload as an AI engineer will be coding. Hands on experience is far more valuable than theory.

Amoy · Answer

You should do CS. At the end of the day companies need someone that can write code and they will filter by degree type. CS will be better.There are also a ton of problems that don’t need deep learning and being able to understand when a traditional algorithm is a better solution will be valuable

AllWillBeWell · Answer

CS degree any day over Statistics degree. Statistics is narrow. CS is broad and will give you access to more jobs than Statistics

maypletree · Answer

I am a ML Eng w/ a grad background in stats (undergrad was a math heavy major) and I would suggest CS + math minor if it's not guaranteed by your unis requirements.Most undergrad stats classes will not be beneficial to you unless you went into data science or a more analyst position. However, classes like probability theory, regression, linear algebra, calculus, and numerical analysis were very helpful. Research experience + learning how to read academic papers is also beneficial.Not having CS as my major made my internship hunt in undergrad a lot harder, and it took me longer to get up to speed. I'm glad I took the path I did, but that was also quite a few years ago when it was easier to break into the field.

ML engineer career advice

About

Tech

Members