AI Engineering Degree Practice Exam 2025 - Free AI Engineering Practice Questions and Study Guide

The statement that a linear relationship must exist between the independent and dependent variables is a defining characteristic of linear regression. This method seeks to find the best-fitting line that describes the relationship between these variables, operating under the assumption that changes in the independent variable(s) will produce proportional changes in the dependent variable.

This assumption is crucial because if the relationship is not linear, the model will likely misrepresent the true nature of the data, leading to poor predictions and insights. Linear regression is fundamentally designed to capture linear dependencies, making this statement true in the context of its requirements and functionality.

In contrast, the other options do not align with the fundamental principles of linear regression. For instance, dependent variables in linear regression must be continuous rather than categorical. Additionally, linear regression can be applied to datasets of various sizes, not just small ones. Finally, while linear regression can be extended or adapted to handle non-linear relationships (through techniques like polynomial regression), standard linear regression itself is not inherently capable of modeling non-linear relationships effectively.