How to excel at Machine learning
ScholarX write up part two
As I have mentioned in one of my previous articles, I was fortunate enough to be selected as one of the ScholarX mentees for the year 2021. My mentor is Dr. Anupiya Nugaliyadda, ACS WA 1962 Medalist for the Most outstanding Ph.D. in 2020. During the last three months, I received a lot of valuable advice as well as insights from him about various aspects of machine learning and AI.
Because I am passionate about sharing what I learn with other learners, I thought I would share the three most important pieces of advice and insights I received with you, especially with the beginners like myself. Hope this will give you some good insights into ML and AI. So without further ado, let’s get into it😋.
Rule no 1: Find what you really want to do. You can’t master all areas in ML
Machine learning is a huge field consisting of computer science, mathematics, as well as statistics. There are numerous areas of ML, such as computer vision, natural language processing, and natural language understanding. It’s not a bad thing to get familiar with a few of those areas, but trying to learn all of them at once can be really hard and frustrating. Once you get familiar with the fundamentals and different study areas in ML, you need to decide what exact area you want to follow and master. That way you can really focus on what you really want and do it. It doesn’t have to be something like computer vision or NLP, it can be something like reinforcement learning that you want to focus on. Or it can be generative adversarial networks that you really want to learn and master. That’s okay as long as you narrow your target down to something you can have 100% focus on.
Rule no 2: Read lots and lots of Research Papers
In the world of ML, new technologies are being invented as well as changed constantly. As a result, keeping up with the latest inventions has become a critical task for an ML practitioner.
And the best way to do that is by reading the papers published by researchers all around the world. Research, like any other scientific field, is what drives ML forward. Reading about them not only keeps you up to date, but also allows you to broaden your knowledge of machine learning. As an ML enthusiast and practitioner, it is therefore recommended that you read research papers as much as you can.
You can find research papers online easily with the Google Scholar web search engine. This search engine is just like google, the only difference is it more focused on finding research papers for you rather than just searching anything on the internet.
Rule no 3: If you feel like giving up, make sure that you don’t
Machine Learning may be an interesting thing to learn as well as practice for many people around the world, But that doesn’t mean it is an easy subject to master. Just having good knowledge of coding is not enough in most cases when you are tackling real-world problems using ML. You need a sound knowledge of mathematics as well statics to do it in a proper way.
Because of the above reasons, I have heard that many people give up learning ML before going deeper into the subject. Even in my short time period where I engaged with ML and DL, I have been hopeless and frustrated countless times. Especially when trying to do research ML. But now I’ve realised one thing:
Frustration comes when you challenge yourself to become something more, not by just trying the same thing again and again.
As my mentor told me, frustration and discouragement are common when attempting to do ML or research (Even the professionals face the same kind of situations). These scientific domains are not meant to be easy. The most import important thing is to keep trying without giving up when you are discouraged or frustrated.
I hope you got something valuable from this article. Please do comment if you have any insights or feedback about the things that I’ve mentioned above. I hope to see you with another article soon. Until then, Goodbye and stay safe!
-Pasan
