Regression

In statistical learning, regression refers to a set of tools that estimate the relationship between dependent (outcome, response, or labels) and independent variables (predictors, covariates, or features).

In general, a key problem with regression models is that they’re sensitive to outliers. Depending on how many there are in the dataset, they can significantly impact the fit line.

Sub-pages

• Techniques
  • Linear regression (simple, multiple)
  • Gaussian process regression
  • Support vector machine
  • Generalised linear model
  • Logistic regression
• Error functions
  • Mean-squared error
  • Root-mean-square error