Bayesian linear regression is a form of linear regression that uses the information about the variances of the input variables (whose errors have to be normally distributed as a precondition to ordinary linear regression anyway) to produce a probability distribution for outputs. While ordinary linear regression predicts a single output value on the basis of a range of input values, Bayesian linear regression predicts a probable value (which will be close to the single output value that the ordinary algorithm would have predicted) and a standard distribution around that probable value.
In situations where the variances of input variables are known prior to training, this information can also be incorporated into the model.
- alias
- subtype
- has functional building block
- FBB_Value prediction
- has input data type
- IDT_Vector of quantitative variables
- has internal model
- INM_Function INM_Probability
- has output data type
- ODT_Quantitative variable
- has learning style
- LST_Supervised
- has parametricity
- PRM_Parametric
- has relevance
- REL_Relevant
- uses
- sometimes supports
- mathematically similar to