The covariance heatmap is added as a measure to asses how well coefficients are correlated in the underlying data. Hereby, a high correlation means that the effect of the coefficients on the output (allecerlation prediciton) is very similar and that the two coefficients can not be observed separatly from each other.
manumerous
commented
3 years ago 
The covariance heatmap is added as a measure to asses how well coefficients are correlated in the underlying data. Hereby, a high correlation means that the effect of the coefficients on the output (allecerlation prediciton) is very similar and that the two coefficients can not be observed separatly from each other.