Simulations using method based on article 'Inference for Regression with Variables Generated by AI or Machine Learning' but generalised to GLM with Poisson distribution.

Contains 25 simulations with 1000 observations each and 25 with 5000 observations each. Similarly as in article, in simulations were used data with measurement error. Results were compared for standard glm ignoring measurement error and implemented method that corrects them.