How often do you use linear models?

I’m curious how consistently people are using linear models in their work. What problems are you using them for? Why are you not using something like a DNN? Do you find it difficult to align your models with stakeholder expectations? How are you implementing these linear models? Python? R?

Would you consider using a calibrated linear model if it meant you could guarantee certain relationships between a feature and the target (e.g. monotonicity) and more granularly understand how your features …

datascience dnn linear people python stakeholder work