I should start my post with writing next sentence: I FINALLY STARTED TO LEARN MACHINE LEARNING.
Yes, I took a course called "Machine Learning for humans" from Ukrainian Data Science teacher Hanna Pylieva.
I enjoy that course and it worth every dollar for all the reasons:
- 🧠 Well-organized course material. Each of next topic based on previous one. There was a topic or the polynomial features I had to return back to Linear Regression to recall about RMSE, MSE, R2.
- 🖥️ Online Zoom meeting with expert Hanna where everyone can ask questions and she will answer!
- 🧑🏫 Curators who help and answer questions of students in Telegram Channel (quite fast answers as for curators I would say).
- 🔥 Of course DEADLINES!! It keeps you motivated and keeps that discipline to learn new and new materials
- 💪 Motivation and support between students. Every one of them is special and keeps that spirit of learning. I even started to recognize some of the students in Hanna's stories 😅
So what about competition??
I just explored Kaggla and found competition from Kaggle on Kaggle (there should be a meme with Obama who gives medal to Obama 😅) called Backpack Prediction Challenge. As for newbie sounds good because data is kinda similar that we did and I decided to use Linear Regression on that competition.
On the Zoom meeting I asked Hanna about that competition - is it good to start and she confirmed that one and even that competition will be counted as practical homework which everyone will do in that course. Sounds good 😏
And you know what? Just in 2 hours after a meeting from 8PM to 10PM I just finished my jupyter notebook with score 39,14866 🎉
After submitting it, I had been read Discussions for 1 hour and found a lot of useful information here like:
- Residual Distribution
- High Noise to Signal ratio
- Feature engineering
Wow... So much information and practical stuff after just one public submit and motivation after Zoom meeting
P.S. ChatGPT and other AI tools were not used to generate that content. All thoughts, and mistakes are mine 😅
