16 key takeaways about starting a career as a machine learning engineer.
↓ 1/10
↓ 1/10
1. There are more courses out there than you'll ever need. Pick one and finish it. There's no such thing as "the best course."
2. You probably already know most of the math you'll need to get started.
↓ 2/10
2. You probably already know most of the math you'll need to get started.
↓ 2/10
3. Don't be discourage by introductory courses that claim you need a ton of math as a prerequisite. Most don't.
4. You want to be proficient with Python. Other languages are great and all, but Python rules.
↓ 3/10
4. You want to be proficient with Python. Other languages are great and all, but Python rules.
↓ 3/10
5. The better developer you are, the easier machine learning will be for you. If you have a hard time programming, it will cost you to make progress.
6. Understanding computer science fundamentals is usually overlooked but extremely beneficial.
↓ 4/10
6. Understanding computer science fundamentals is usually overlooked but extremely beneficial.
↓ 4/10
7. While learning, don't worry about the code and invest your time understanding how and why things work.
8. Some algorithms are complex as shit. 99 out of 100 times, you can just use them without worrying about how they work.
↓ 5/10
8. Some algorithms are complex as shit. 99 out of 100 times, you can just use them without worrying about how they work.
↓ 5/10
9. "Understand every detail of an algorithm if you want to use it" is bullshit advice.
10. Machine learning is not different than any other subject you may learn. You can start at a very high level and get deeper into details as you need them.
↓ 6/10
10. Machine learning is not different than any other subject you may learn. You can start at a very high level and get deeper into details as you need them.
↓ 6/10
11. There are 375,475,484 algorithms out there. In practice, you'll probably use a handful of them.
12. You don't need a GPU to get started. You don't need a GPU most of the time. You will at some point for deep learning, and you can solve it online for free.
↓ 7/10
12. You don't need a GPU to get started. You don't need a GPU most of the time. You will at some point for deep learning, and you can solve it online for free.
↓ 7/10
13. Most machine learning educative learning materials end when you finish training your model. In real life, this is where the fun starts.
14. The best solution for a problem is not necessarily machine learning-related. In fact, this is the case most of the time.
↓ 8/10
14. The best solution for a problem is not necessarily machine learning-related. In fact, this is the case most of the time.
↓ 8/10
15. The loudest machine learning voices right now come from academia. Make sure you understand that a research position is very different than working in the industry.
16. Pragmatism is a great trait when building a career centered on providing value.
↓ 9/10
16. Pragmatism is a great trait when building a career centered on providing value.
↓ 9/10
If you are looking for a pragmatic perspective on building machine learning systems, follow me @svpino and subscribe to my newsletter digest.underfitted.io.
It's free, and you get one story every Friday right in your inbox.
10/10
It's free, and you get one story every Friday right in your inbox.
10/10
Absolutely! A lot of these things are the same across different disciplines.
https://twitter.com/s_adamiak/status/1379815308689616897?s=20
If you tune out all of the noise, you can focus on delivering value.
https://twitter.com/ggospodi/status/1379884528324448256?s=20
• • •
Missing some Tweet in this thread? You can try to
force a refresh
