by Erika Rykun, Udemy 3 February 2020
Artificial intelligence is heralded as possibly the most disruptive technology in history, as AI can be implemented in so many industries. Machine learning is what drives AI development, and its subset deep learning is taking us even closer to general artificial intelligence.
With so many possibilities for AI, it’s an opportune time to start learning the fundamentals of machine learning. In this article, we’re going to highlight five ways you can easily get up to speed on machine learning, particularly for DIYers. Just please don’t go creating Skynet with your newfound skills.
Brush up on your math and programming skills
An understanding of machine learning requires at least a basic understanding of linear algebra, matrix algebra, and calculus. The field of machine learning draws techniques from aspects of statistics, data science, probabilistic logic and computer science. Your level of understanding machine learning can directly correlate to your understanding of various math topics.
As for the programming aspect, most machine learning is done in Python, a language which you’ll certainly want to learn as well. Once you understand Python, you’ll want a more defined focus in Python for data science applications.
Join online communities
Knowing where people learned in machine learning congregate will put you in contact with experts in the field. On communities like R/LearnMachineLearning on Reddit, you can get insights by browsing topics related to machine learning, or post questions you have about ML. You’ll also see a lot of great posts and links related to research papers and lab projects.
Aside from ML-specific communities, it would help to join other coding communities, such as those specifically for Python, or more general mathematical coding on StackExchange.
For live chat rooms rather than forum-style resources, you can discover a wealth of professional and academic servers to join. Here is a great list of data science servers you can join in Discord.
Enroll in online courses
While the majority of professional programmers entering the workforce today are self-taught, machine learning is one field where education will really help. Online courses can offer self-learning courses, virtual classrooms, or a combination of both, so you really have your options depending on how you prefer to learn.
By taking a structured course, you’ll be introduced to concepts and techniques you may not have discovered on your own, and certification certainly doesn’t hurt job prospects if the goal is a career, rather than a hobby. We could list some online courses, but it’s best if you research one that is best suited for you.
Some courses are entirely free and may offer a more general overview of machine learning, while paid courses may focus on more specific applications of machine learning, and offer certification.
Watch free online videos
Videos are often better than textbooks, and websites like YouTube are a veritable treasure trove of machine learning content. You can just type “introduction to machine learning” and get tons of videos, from live courses to machine learning tutorials and follow-along exercises. There are really very few things you can’t learn on YouTube.
Tinker with Github code
When you have a more proper understanding of machine learning, you can begin tinkering with code. Github is the world’s leading software development platform and code repository with many open-source projects you can clone and experiment with. Github hosts some great material on machine learning topics as well, so overall it’s an excellent site for your purposes.
This will be by far the best method of learning machine learning, once you reach this stage, as you’ll be able to put all the knowledge you’ve gained to real, applicable use.