Types of Regression Lines
1. Linear Regression Line: Linear regression line is utilised when there is a linear relationship between the reliant variable and at least one free variables. The condition of a straightforward linear relapse line is typically; Y = a + bX + ε, where Y is the reliant variable, X is the free variable, a is the y-intercept, b is the slope, and ε is error.
2. Logistic Regression Line: Logistic regression is used when the dependent variable is discrete. It models the probability of a binary outcome using a logistic function. The equation is typically expressed as the log-odds of the probability.
3. Polynomial Regression Line: Polynomial regression is used when the relationship between the dependent and independent variables is best represented by a polynomial equation. The equation is Y = aX2 + bX + c, or even higher-order polynomial equations.
4. Ridge and Lasso Regression: These are used for regularisation in linear regression. Ridge and Lasso add penalty terms to the linear regression equation to prevent overfitting and perform feature selection.
5. Non-Linear Regression Line: For situations where the relationships between variables is not linear, non-linear regression lines must be used to defined the relationship.
6. Multiple Regression Line: This involves multiple independant variables to predict a dependant variable. It is an extension of linear regression.
7. Exponential Regression Line: Exponential Regression Line is formed when the data follows an exponential growth or decay pattern. It is often seen in fields like biology, finance, and physics.
8. Pricewise Regression Line: In this approach, the data is divided into segments, and a different linear or no linear model is applied to each segment.
9. Time Series Regression Line: This approach is used to deal with time-series data, and models how the dependent variable changes over time.
10. Power Regression Line: This type of regression line is used when one variable increases at a power of another. It can be applied to situations where exponential growth does not fit.