Solve a thousand problems
From data-driven algorithm based on artificial intelligence to data-driven based human intelligence (SWE)
If you are learning machine learning, you may have a high possibility to hear “Data-driven algorithms”. As a human being with limited intelligence, I am trying to learn something from a computer. So I am using a thousand problems to train myself and hopefully I can do well on testing (Job interview).
What do you mean about a thousand problems?
Yes, it is a thousand Leetcode problems. As you can see here.
About two years ago, I have solved 200 questions. For the last whole year, I solved another 800 questions.
Here are some quick conclusions:
- Training can help you do well on regular coding interviews: Yes, I did pretty well on the coding interview in terms of algorithm and data structure If the questions are not that hard. I still have a good chance to fail the coding interviews from top companies especially during this pandemic season.
- Timing is pretty important: I failed most of my interviews when I can solve about 500 problems. However, I got my internship from Google even I only solved 200 problems. At the same time, I got some offer after I solved about 800 problems, 60% of the reason due to my proficiency in coding.
- Training can build my confidence: keep on practicing and doing well on coding questions can build up my confidence which is pretty helpful for interviews.
- Training on coding only is not all you needed: you also need to know some other things such as OOP, behavior questions. If you are looking for MLE, you also need to know Machine Learning.
An important observation
When I tried to follow the computer’s way to build up myself, I find a significant difference between myself and the computer. During the data-driven process, I can see the beauty of the algorithms and codes. Maybe I should say elegant. I am not clear about the feeling of computers during their training process. Maybe what they can feel is only hot as CPU/GPU are keeping on running. Maybe they have some other feelings. We will see.
Thanks * INF
Thanks to so many people’s help during this process. Although, lots of help is from the videos or online materials you guys shared on the internet. The appreciate list is not limited to those people listed HERE.
Thanks for your time on my random talk. Hope you don’t dislike it. If so, clap it let me know you somehow feel this article is a little bit helpful. Thanks.