Tolerance (TOL) and Variance Inflation Factor (VIF) are used to detect the multicollinearity among your predictor variables (Independent variables). Detecting multicollinearity is important for hazard prediction mapping because highly correlated variables can be a noise in the model and even reduce the accuracy of the final model. By doing simple linear regression you can find the VIF and TOL values. According to the present literature a Tolerance value less than 0.20 or a VIF of 5 and above are treated as multicollinearity.
Achu A .L hi, Achu, could you please tell me the reference of your sentence ( According to the present literature a Tolerance value less than 0.20 or a VIF of 5 and above are treated as multicollinearity). Because i will need more details and support reference about this when conducting multi-regression model for prediction.