MrudhuhasM / Regression

A simple notebook explaining simple linear regression

Regression

In this repositery i am going to upload few notebboks on different types on Regression analysis

• Simple linear regression
• Multiple linear regression
• Polynomial regression
• SVM Regression
• Decission tree regression
• Random Forest regression

Linear regression have few assumptions.Before proceeding forward we need to make sure the data is met with assumptions

1. Linearity : There should be a linear relation between independent varibles and dependent varibles
2. Homoscedaticity : Residuals should be normally distrubuted and should have equal variance
3. No Endogeneity : Residuals and independent variables should be correlated
4. No Autocorrelation : There should be no correlation in residual terms
5. No MultiCollinearity : Independent varibles should not be correlated

If the assumptions are not met we could transform the data or we could use non-linear regression model

Linear regressions

• Simple linear regression
• Multiple linear regression

Non-linear regression

• Polynomial regression
• SVM regression
• Decission forest regression
• Random forest regression