Categories of algorithms
We've categorized ML algorithms by their learning mode, but that's not the only way to categorize algorithms. Another approach is to categorize them by their task or function. In this section we will briefly present the basic functions of ML algorithms and name some example algorithms.
Clustering
Clustering algorithms aim to identify groups of data points that are similar to one another. The definition of similar depends on the type of data, the problem domain, and the algorithm used. The easiest way to intuitively understand clustering algorithms is to visualize points on an x/y grid. A clustering algorithm's aim is typically to draw circles around groups of similar points; each set of circled points is taken to be a cluster. The clusters are generally not known beforehand, so clustering algorithms are generally classified as unsupervised learning problems.
Some examples of clustering algorithms include:
- k-means, and variants such as k-medians
- Gaussian mixture models...
