Big Data and Machine Learning

Abstract

This chapter focusses on the opportunities and challenges of big data and machine learning in relation to oral epidemiology. Big data are characterized as high-volume data that allow for smart data processing and integration of multiple data sources. They can be useful in many ways for studying the distribution and determinants of oral health. Machine learning, particularly the development of predictive models using artificial intelligence, offers great potential for the improvement of people’s oral health and care through the exploitation of big data. The knowledge derived from big data analytics can help improve oral health policy and clinical decision-making by better specification of intervention points. However, care should be applied regarding potential data quality threads such as data entry errors or non-harmonized standards for data coding. Unless based on sound theoretical frameworks and appropriate statistical methods, any type of big data analysis is prone to be corrupted by spurious correlations and fallacious inference. If used judiciously, however, big data offer vast opportunities for the creation of knowledge and (artificial) intelligence for the better promotion, protection, and management of people’s oral health.