A Machine Learning Specialist works for a credit card processing company and needs to predict which transactions may be fraudulent in near-real time. Specifically, the Specialist must train a model…

Streaming classification refers to the operational aspect of applying the model to a continuous stream of data, not the fundamental type of machine learning problem being solved.

This business problem is framed as binary classification because the goal is to categorize each transaction into one of two mutually exclusive classes: "fraudulent" or "not fraudulent." Additionally, predicting the probability of fraud is a common output of binary classification models, allowing for risk-based decision making.

Multi-category classification involves predicting one of three or more distinct categories, whereas this problem only has two outcomes (fraudulent/not fraudulent).

Regression is used for predicting a continuous numerical value, such as house prices or temperature, not for categorizing items into discrete classes.