How I Got Into Machine Learning and Python – In the beginning, I wasn’t even interested in computers and new technologies. I was only interested in becoming a national champion in a ridiculous game called “Icy Tower”. Wait, what? Yes, you heard me right.
So it all began from “Icy Tower”… How absurd is this? I’ll let you decide. Basically when I was a kiddo, it didn’t take much for me to get hooked to the computer for hundreds of hours, and when I think about that, after 20 years, nothing has really changed. Only the purpose evolved constantly and that’s how my interest went from “Icy Tower” (which isn’t the most compelling game created by human beings), to chess, then playing professional poker, and finally getting hooked on Machine Learning.
When I got to the point when my previous activities weren’t giving me as much joy as I expected, I started looking for something else and that’s how I found programming, and subsequently Machine Learning. The first choice I had to make was picking the right programming language. There were only two options: Python and R. I wanted to have a wider range of possibilities so Python was the obvious choice.
At this time I was in the zone, and I could’ve read about the topic all day long without even a small break. I didn’t want to stop the obsessive part of me because it was pure joy. After a couple of weeks, I had a solid general understanding of everything connected to Machine Learning. And guess what?
As smart people suggest, I started learning linear algebra, analytic geometry and vector calculus. Not. Like most people, from here, I jumped straight to the top of the iceberg, which for me was Reinforcement Learning. Believe me, it was a pretty big leap. So without much foundation, I decided to choose one of the hardest parts of Machine Learning.
Yeah, another ridiculous choice that turned out to be brilliant after a couple of months. But you might be wondering… why? And the answer isn’t very obvious because everything stands against this decision. I wasn’t aware of how much effort I had to put into plugging all the leaks and learn everything with an acceptable level of understanding. The only thing that kept me motivated was my obsessive interest. If I could give you one piece of advice, it would be this: Choose the topic that interests you the most.
Nothing else matters. There will be many moments of despair and what helps you overcome them, is the joy coming from doing what you’re passionate about. After 2 or 3 months of full-time learning, I became quite familiar with Reinforcement Learning, just to discover there are no jobs in this field. There were many practical topics to explore, so I decided to commit to self-learning full-time again and there was no better way to master Machine Learning, than through working on real-world projects.
It took about 6 months of mastering different areas of Machine Learning like NLP, Computer Vision, tabular data regression and classification, databases, model deployment, and so on, before I felt I was ready for the real projects. The process of getting the first end-to-end project experience took another 2 months of sending multiple applications every couple of days.
The more knowledge and experience I had, the more tempting it was to share it with the people around me. I found it really rewarding to see how I can help other people in their Machine Learning development. That’s priceless, and it also keeps me motivated to be a better version of myself every single day. I hope all of you will stay involved in projects you truly love.
reference – How I Got Into Machine Learning and Python
Web enthusiast. Thinker. Evil coffeeaholic. Food specialist. Reader. Twitter fanatic. Music maven. AI and Machine Learning!