Showcasing calculations for multiple linear regression, based on first principles - with more than 1 independent variables
Performing Linear Regression from First Principles
- Performing checks on linear regression assumptions of linearity, normality, homoscedasticity, indepedence and multicollinearity
- Data Transformations
- Calculation of Correlation Coefficient between variables from first principles
- Obtaining solution to linear regression
- Standard error of the regression (model), mean, and coefficients
- R-squared and adjusted R-squared
- Comparison of calculations with open-source libraries