When we first venture into the use of artificial intelligence for data analysis, the first problem we are faced with is to choose the most appropriate algorithm for solving a specific problem. Analyzing the available algorithms, we immediately realize that the choice is not so immediate and requires an appropriate investigation.

A first approach to the problem involves the specification of the task that our machine learning algorithm will have to face. In this sense we can rest assured: there are only a handful of tasks to be analyzed even if, for each of these activities, different approaches and algorithms are available.

In fact, even if all machine learning algorithms take the same data as input, what they'll want to achieve is different. Machine learning algorithms can generally be classified into a few groups based on the tasks they were designed to solve. The typical activities in any automatic learning are as follows:

• Regression
• Classification
• Clustering
• Dimensionality reduction

In the following sections, we will analyze the characteristics using real-life examples.