I am a thermal power engineer, graduating with a master's degree in 2019. I have been working as a data analyst for the last year. I am the only analyst in the department, mostly doing ad-hoc tasks, mostly using SQL and Python. Before that, I worked at the intersection of thermal power and IT for 2,5 years.
I originally wanted to gain knowledge in machine learning. After the February events I came to the conclusion that it was time to change my profession to the appropriate one.
I liked the first two blocks very much, I enjoyed the course for all 4 months. Somewhere from the third or fourth lesson of the third module the tasks became impossible for me. Since I was very responsible in my approach to the course, I had a kind of burnout because of it.
Why the tasks stopped being feasible is a multifactorial phenomenon: the complexity of the topic, burnout, and a drop in the quality of the lecture material. Judging by the survey in Telegram, the majority also liked the first two modules, and there were many more questions for the third and fourth modules. The quality of the lectures in the fourth module raised the most questions: 20-minute videos with an average duration of an hour and a half for the rest of the course, little practice, just definitions and sometimes screenshots of solved problems.
Nevertheless, the tasks were quite doable, but I can't say that I got anything out of this module in my head. So a special thanks for opening a block on statistics from the Data Analyst course. But I haven't touched it yet because of the accumulated fatigue from studying.
The final project is cool in terms of stack and idea, but it was too big and without hints (as I realised, they appeared on the next threads). Because of that, only a few people turned the project in before the deadline, as I understand it. Still, it's what the course is all about.
The fifth block on interviews is very interesting due to algorithms, special thanks to Lyosha Kozharin for these 4 lessons. Unfortunately, I didn't hand in all the tasks on time, as my "burnout" hadn't passed by that time :(
I think Docker should be added to the first module. But as I found out recently, K/C has a free course on Docker planned, which sounds awesome. I would also add more information on bousting in the second module.
The course hasn't impacted my career yet, but hopefully it will. I'm not sure what to mark as results - getting familiar with ML basics really went well, repeating and strengthening the Python base too.
I hope that the third module will become simpler. For example, instead of one huge task, there will be several smaller tasks, as in the second module. In the fourth module, we should double (or even triple) the timing of lectures and add as much practice not with screenshots of written code, but with live coding and comments.
I would like to thank the whole team of course creators for quite a high-quality course on the basics of machine learning. I don't have much to compare it with, a year ago I took a course from Learn Python, but it was very basic. I didn't know about K/C at that time, otherwise I would have chosen a course on data analytics.
Perhaps we should reward the most active students somehow - for example, give a ticket to Matemarketing, give a big discount on another course or something nice.
David T. Wood
START ML