Unsupervised learning

Similarly, in the case of unsupervised learning, there is no target attribute. The objective of unsupervised learning is to identify patterns by deducing structures and the relations of the features in the input dataset. It can be used to discover rules that collectively define a group, such as topic generation, partitioning—such as customer segmentation or determining the internal structure of the data such as gene clustering. Examples of unsupervised learning algorithms include association rule mining and clustering algorithms.

It is quite essential to know about different learning algorithms before creating an AutoML system. Before using an algorithm, it is critical to understand its triple W—What it is, Where is it used, and by What method it can be implemented.

In the following sections, we will question different algorithms for their triple W, which will aid in creating a robust AutoML system.