Machine Learning in Official Statistics: A Preface-Like Introduction
摘要
Machine learning. What is that supposed to be? Do we need it? Isn’t this just something that ‘they up there’ heard about and now want to have too? Or is it perhaps something that computer nerds claim to need, even though it’s just a gimmick? These or something similar were probably the questions in many statistical offices when the first attempts were made to take on machine learning 10–15 years ago. The early adopters certainly did not do this by first designing a strategy and then setting up a project with a steering committee and controlling formal matrix structure.