Why revisit K-Means?
This tutorial is very brief as you can see. The only purpose of this tutorial is to discuss why we need to revisit K-means as we delve deeper into SOMs.
We discussed K-means before in one of our previous machine learning courses. Our next tutorial in this section will be based on a video taken from that course.
That tutorial will be an introduction to K-mean clustering to those of you who haven’t taken the course and memory refreshment for those who have.
Learning about K-mean clustering will be extremely helpful when dealing with self-organizing maps. It’s not a major part of SOMs, but it will prepare you to understand them properly.
There are two major elements in K-means that will provide you with that introduction to both SOMs and unsupervised machine learning in general:
- You will get to see in the next video the push-and-pull movements that these nodes go through as they travel across the map. That process will give you a glimpse of what you will deal with when working with SOMs.
- Like SOMs, K-means are also unsupervised, although the K-means method is merely a machine learning algorithm rather than a neural network.
That’s it for this tutorial. You can now move on to the actual K-means tutorial.