Tuesday, October 26, 2021

Multiple_linear_regression

 # What is multiple_linear_regression

* Multiple linear regression (MLR), also known simply as multiple regression, is a statistical technique that uses several explanatory variables to predict the outcome of a response variable.


* Multiple regression is an extension of linear (OLS) regression that uses just one explanatory variable.


* MLR is used extensively in econometrics and financial inference.


* Regression models are used to describe relationships between variables by fitting a line to the observed data. Regression allows you to estimate how the dependent variable changes as the independent variable(s) change.


* Multiple linear regression is used to estimate the relationship between two or more independent variables and one dependent variable. You can use multiple linear regression when you want to know:


* How strong the relationship is between two or more independent variables and one dependent variable (e.g. how rainfall, temperature, and amount of fertilizer added affect crop growth).


* The value of the dependent variable at a certain value of the independent variables (e.g. the expected yield of a crop at certain levels of rainfall, temperature, and fertilizer addition).


*---------------------------------------------------------------------------------------------------------


* Linear regression is one to one but Multiple regression is many to one relationship

* Adding more independent variables to the model doesn't mean that regression will give better in fact it will be worse. This can be overfitting 

* Adding more independent variables will lead to multicollinearity.

* The idea of all independent variables to be co-related with the dependent variables,s not with each other.


Python code is available on GITHUB 

GITHUBLINK

Multiple Examples of Multiple Linear regression  are available on below link:

GITHUB


No comments:

Post a Comment

"🚀 Delta Lake's Vectorized Delete: The Secret to 10x Faster Data Operations!"

"🚀 Delta Lake's Vectorized Delete: The Secret to 10x Faster Data Operations!" Big news for data engineers! Delta Lake 2.0+ in...