Understanding Machine Learning Concepts Using Visual Aids
A purely mathematical explanation of machine learning concepts is not for everyone. If you don’t have the appropriate background it might even rather confuse than help you. Personally, I always had a hard time understanding clustering and dimensionality reduction tools. In order to understand these concepts, I heavily relied on visual explanations.
In this article, I put together a list of my favorite interactive visual explanations of machine learning and deep learning techniques currently on the internet (that I know of). I hope this will help you understand these topics better, give you an intuition of how these algorithms really work and therefore help you make better decisions.
Principal Component Analysis
Victor Powell and Lewis Lehe have an amazing blog with numerous visual explanations of a plethora of topics, from Markov chains to conditional probability. One of my favorite articles is the one on Principal Component Analysis. It is interactive, visually pleasing and there are real-world examples.
Decision Trees
If you are new to machine learning, I highly recommend you to look at this article. While it explains the basic concept of decision trees, it is also an amazing primer for the field as a whole. Plus, the article is available in 13 languages!
The Bias-Variance Tradeoff
This is a fundamental topic in machine learning, yet every time I have to explain it to someone, I struggle. This article helped me greatly to get a deeper understanding of it and its practical implications. It is the follow-up article to Decision Trees, so I highly recommend to look at these two in conjunction.
t-SNE
As I mentioned above, I always had some issues with dimensionality reduction techniques. This article will teach you that hyperparameters do matter, that randomness has many faces and much more.
Gaussian Processes
Have you ever wondered what the heck gaussian processes are? This article can help you get a proper intuition of this concept. While it is a bit math-heavy, knowing about gaussian processes will open your eyes to a whole new world of machine learning.
Neural Networks
What happens when I add a layer to my neural network? If I change activation functions? When I change batch size? This is an amazing interactive tool to visualize neural networks, brought to you by Tensorflow.
Gradient Boosting
In my experience, Gradient Boosting algorithms are applied but not really understood. While most have a conceptual understanding of Random Forests, Gradient Boosting is just seen as its extension. But is it really? Check out this article for an amazing visual explanation of these algorithms.
Wrap Up
I hope these links will help you get a deeper understanding of some of the most fundamental concepts in machine learning.
Please let me know if you know other similar articles.
Lastly, I want to thank the authors for their commitment to teaching everyone on the Internet for free!
