Well I'm back again (that's a good start to my 100 days of learning machine learning).
Day 2
I moved onto the next chapter of the scikit-learn data camp course. Today was all about linear regression. I can really see what people mean that it only takes a few lines of code to do machine learning. Both KNN (see yesterday's post) and linear regression use like 10 lines top. So that's nice.
Anyway, today I managed to get a linear regression model working which I'm pretty happy about. I also now know what ridge regression and lasso regression is. They were always just mystery words that people would throw out sometimes. Like if you order a a plain black coffee and then they say do you want sugar or milk with that. I'd just look at them in confusion and be like "I just wanted coffee??". But now I understand that lasso and ridge regression are just ways to supplement linear regression if you need them. Just like milk and sugar with coffee (I think, I don't actually drink coffee)
I'm unsure what my next step should be. I could carry on with the course and learn a bit more about hyper parameters or I could jump over to kaggle and try building my first model. I want to do both so it's just a question of which one I do first. What do you think I should do?
Be back on Monday for day 3, have a good weekend!
Top comments (0)