R codes for performing Regression analysis
split the given data into train and validation data sets
calculated residuals (Studentized residuals) performed outlier analysis using various criteria like i) cook's distance ii) DF BETAS iii) COV ratio
checked for multicollinearity from cov matrix
result no multicollinearity
ploted ridge plot
After we fit the model, number of significant regressors are A1,A4,A2 and intercept
overall model is significant and in case of individual significance A1,A2,A4
This means octane rating depends on amount of material 1,2 and 4
summary(model) tells that they are 4 variables and 44 degrees of freedom
sum of squares of residuals( residual standard error) is 0.4395