top of page


Regression displays the relationship between the input and output. This is one of the biggest parts of machine learning. There are multiple types of regression and all will be explained. The major types though, logistic and linear regression, will be covered in depth.

Types of Regression

  • Linear Regression - 

  • Logistic Regression - 

  • Ridge Regression - This model is used to analyze multiple types of regression (linear, exponential, etc.) data and the minimized cost is unbiased, but are extremely variated across weights. This model tries to minimize the variance between the different costs. 

  • Lasso Regression - Lasso regression is a supervised machine learning method to constrain values so that the regression coefficients (i.e. weights) for a few variable shrink towards zero. This allows the model to identify the correct weights at a greater accuracy. 

  • Polynomial Regression - This type of regression fits a wide range of inputs to help model the relationship between the input and output best (the polynomial version of linear regression).

  • Bayesian Linear Regression - This is a type of linear regression that uses probability distributions (rather than the typical point estimates used in linear regression). The output is drawn from probability distributions, rather than being a single value.

bottom of page