From Displayr
Jump to navigation Jump to search

Regression quantifies the relationship between an outcome variable and one or more predictor variables. It is primarily used to understand causal relationships. For example, how will sales increase if price is dropped by $1? How is life expectancy changed by consumption of a new drug?

The above definition is very broad and encapsulates much of machine learning. However, most of the time when people refer to regression they are referring to Linear Regression.

Creating regression models in Displayr

Various models are available via Anything > Advanced Analysis > Regression.

Alternative meanings of "regression" in data science

There are two other meanings of regression within data science:

  • Regression tests and regression testing refers to the process of ensuring that software gives consistent results over time. This is the main usage of regression within software engineering.
  • Regression can refer to the tendency of data or behaviors to revert over time to some typical or base level (e.g., regression to the mean).