Applying classification techniques
Classification differs from regression, where the goal of the classification technique is to assign each data point one or more predefined labels based on the variables. While regression predictions can theoretically output any continuous number, classification outputs are predefined from the start. For example, classification might produce outputs such as Spam or Not Spam in email classification scenarios.
Three common classification cases exist, including the following:
- Binary classification: There are only two possible classes for the output, where the model decides between a positive and a negative outcome
- Multi-class classification: There are more than two classes for the output, but the data only belongs to precisely one of them
- Multi-label classification: The data could simultaneously belong to several classes (labels are not mutually exclusive)
Classification is essential to learn because it is one of the...
