The ML Script : Week 3

Hello world!

The past week was very lazy and not very productive. But I maintained my word of giving at least 45 minutes per day to ML, so yeah, there is that. Long story short, some stuff happened to me—or I intentionally let it happen—and my motivation lowered drastically, but I had committed to 45 minutes every day, so the streak is on. It’s just that my average learning per week has decreased significantly. I am trying to see the problem and address it. Well, enough crying from my side.

Last week, I planned to learn the intuition behind Gradient Descent. I tried to understand it and its types. Gradient Descent looks kinda weird and different from the regular math we studied back in school or here in college. It is like a hit-and-trial for a specific number of times (called epochs).

Aside from Gradient Descent, I tried to implement the mathematical intuition behind it with code, and I failed miserably. It is way different from the other method, i.e., OLS (Ordinary Least Squares). Anyways, it’s fine; I may have to look at it again in the future. But let’s see.

After that, I was reading about data leakage, cardinality, and some evaluation techniques.

So yeah, I guess that’s all I did. Not much, I know. But who cares (except for me)? I will try to do it—well, I am doing it already, I just need some efficiency in the job. So yeah, see you next week. If you don’t find a post, then assume I am slacking off.

Till then…

Huh!

Who cares?

Do whatever you want.