From technology to methodology: A machine learning framework for high quality statistics
Conference
65th ISI World Statistics Congress 2025
Format: IPS Abstract - WSC 2025
Keywords: artificial intelligence, machine learning, methodology
Session: IPS 1005 - Building Bridges Between (Official) Statistics and Machine Learning Methodology
Wednesday 8 October 10:50 a.m. - 12:30 p.m. (Europe/Amsterdam)
Abstract
The rapid development of machine learning as a technological tool has led to the development of new algorithms and applications at an unprecedented rate. Despite rapid technological progress, the field has often been unable to address critical statistical issues necessary to produce high-quality, reliable results. This presentation examines the integration of technological developments into more rigorous methodologies by presenting a comprehensive framework for accounting for all sources of error in machine learning models, the 'Total Machine Learning Error Model,' that provides insight into how to overcome these shortcomings. With this framework, technological advancements will be matched with methodological rigor by improving machine learning models and machine learning-based statistics.