This article is based on the experiences of machine learning engineer Daniel Burke:
I am a machine learning engineer and what I have written for you are tips that will be very useful and motivating for beginners. If I start machine learning again, I’ll make a note of these and be sure to implement them all.
I remember many years ago in Australia we hosted a robotics meeting and answered the questions of the participants. Someone asked a question: How can I enter artificial intelligence and machine learning from a different field without any background?
My colleague Nick came back and called my name. “Where’s Dan Burke?” I was backstage talking to Alex. Nick is the CEO and founder of Max Kelsen, a technology company in Brisbane.
Nick continued: “Of course. she is here.” Dan majored in health sciences, majored in nutrition, and then worked at Uber for a while. He learned machine learning online and has been working with Max Kelsen as an expert machine learning engineer for a year now. If you have any questions, I’m sure he’ll be happy to help you.” And I answered his questions.
Another person, Enkit, talked about a project he was working on to better understand his students’ learning. He was combining lecture attendance rates, time spent on the online learning portal, test results, and a few more. He has even created a front-end web portal to interact with the results. His work inspired me.
That day many other people started coming and asking questions about how to get into machine learning. They were all from different disciplines. Since I am sure that every day many beginners who work in different fields want to enter the world of machine learning, I felt it necessary to share some of my experiences with you.
One of my greatest experiences, which I also recommended to my brother who is starting to learn to machine, is that the best coach is someone who is 1-2 years ahead of you. That is, someone who has just gone through what you are about to go through. You can easily learn new things from him in the best way.
Table of Contents
My most important experiences regarding machine learning training
a) Learn some basics of Python (required time 3-4 months)
Artificial intelligence and machine learning are tied to the Python language. Of course, before Python, you may start R, Java, or any other language. What is important is to choose one of the languages and stick to it! If you want to get into applied machine learning, learning Python coding is a must.
Choose a basic course through the Internet and follow the training with determination and continuously for a few months. You don’t need to become a professional in that language. Of course, you will get more points when you also teach data science. DataCamp is perfect for this. Although it will be difficult at times, my suggestion is to learn a programming language like Python in line with Data Science at the same time.
b) Start when you’re not ready
Use what you have learned as soon as possible. It doesn’t matter how many courses you have taken. Of course, you will never be 100% ready. In any case, to create a professional resume and demonstrate your ability, do not go through many training courses without starting a job. This is one thing I would change if I were to go back and start over. To learn more, just find your project to work on and learn by trial and error.
Consider the speech of a 3-year-old child. He expresses the words constantly and conveys his meaning. He doesn’t have the correct pronunciation, and he doesn’t express the sentence structure or the grammar correctly, but in the end, he will learn the mother tongue completely.
c) There are many resources for learning, be careful not to get confused
Many courses on machine learning are held every day. Most of them are usually excellent and it is rare to find a completely worthless course. However since there are many sources, it is difficult to choose them for training machine learning. This is another pitfall that can hold you back. To further research and delve into this field, I even got my Master’s degree in AI. No one can determine which resource will be suitable for your learning better than you.
But as the best resources, I’d recommend the three that most closely match what I do every day: Hands-On Machine Learning Books, fastai’s Machine Learning Course, and the Applied Data Science with Python course on Coursera. So it is better to remember these things and use them after a few months of experience working with Python.
d) Pursuing research is necessary, but if you can’t use it, it’s useless
You have probably seen a lot of articles and research published every day about new machine-learning methods. Ignore them. There is no way to use all of them and it will only cause disruption and disorder in your learning path. Many of the best machine-learning techniques have been around for decades. What has changed is the increase in computing power and data availability. So don’t get distracted by new methods. If you’re just starting, start by strengthening your foundations. Then expand and update your knowledge as your project requires.
e) Practice a little every day
At the age of 3, you started learning your mother tongue. Within two years, you went from not understanding a single word to being able to converse with people, like your parents, who had been speaking the language for years. But how? Because you practiced every day unintentionally. After a while, you became curious to learn more. If you become 1% better than the previous day in learning a skill, you will be 360% better at the end of the year. If you are tired and bored one day, it doesn’t matter because it is normal and it happens to everyone. The important thing is to resume learning when it is possible again.
f) Do not blame yourself for not knowing something
Have you ever built a recommendation engine? “No”! We have a project that needs this engine, do you think you can write it? “Sure, yes.”
I have never made a suggestion engine either. But I did it in a week. I learned a lot of new things while doing it. Most people think that learning stops after high school or college. They are completely wrong.
It’s okay if you fail or fail at something you’ve already done. You’ve walked all your life, but you don’t blame yourself when you fall. This is normal for you and you continue walking. So why is it so difficult to accept uncertainty and failure in a new experience that you have never had the experience of doing before and that is completely new to you?