When linear regression assumptions are broken, least squares regression is ineffective. Examples are listed below
Non-Linearity: If the independent and dependent variables have a non-linear connection, least squares will fail to capture it.
Heteroscedasticity: When the variance of errors varies between data, least squares might produce wasteful results.
Autocorrelation: If residuals are correlated, as is common in time series data, the least squares assumption is broken.
Outliers: The presence of outliers can skew least squares estimates, resulting in biassed findings.
Multicollinearity: When independent variables are highly correlated, the variances of coefficient estimates increase, rendering them untrustworthy. for Demo Online Class Contact us Durga Online Trainer