{"title":"Predicting and explaining with machine learning models: Social science as a touchstone","authors":"Oliver Buchholz , Thomas Grote","doi":"10.1016/j.shpsa.2023.10.004","DOIUrl":null,"url":null,"abstract":"<div><p>Machine learning (ML) models recently led to major breakthroughs in predictive tasks in the natural sciences. Yet their benefits for the social sciences are less evident, as even high-profile studies on the prediction of life trajectories have shown to be largely unsuccessful – at least when measured in traditional criteria of scientific success. This paper tries to shed light on this remarkable performance gap. Comparing two social science case studies to a paradigm example from the natural sciences, we argue that, in addition to explanation, prediction is an important goal of social science – and we identify constraints that impede pure ML prediction from being successful in that field. As a remedy, we outline elements of an integrative modelling approach that combines explanatory models and predictive ML models.</p></div>","PeriodicalId":49467,"journal":{"name":"Studies in History and Philosophy of Science","volume":null,"pages":null},"PeriodicalIF":1.4000,"publicationDate":"2023-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Studies in History and Philosophy of Science","FirstCategoryId":"98","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0039368123001425","RegionNum":2,"RegionCategory":"哲学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"HISTORY & PHILOSOPHY OF SCIENCE","Score":null,"Total":0}
引用次数: 0
Abstract
Machine learning (ML) models recently led to major breakthroughs in predictive tasks in the natural sciences. Yet their benefits for the social sciences are less evident, as even high-profile studies on the prediction of life trajectories have shown to be largely unsuccessful – at least when measured in traditional criteria of scientific success. This paper tries to shed light on this remarkable performance gap. Comparing two social science case studies to a paradigm example from the natural sciences, we argue that, in addition to explanation, prediction is an important goal of social science – and we identify constraints that impede pure ML prediction from being successful in that field. As a remedy, we outline elements of an integrative modelling approach that combines explanatory models and predictive ML models.
期刊介绍:
Studies in History and Philosophy of Science is devoted to the integrated study of the history, philosophy and sociology of the sciences. The editors encourage contributions both in the long-established areas of the history of the sciences and the philosophy of the sciences and in the topical areas of historiography of the sciences, the sciences in relation to gender, culture and society and the sciences in relation to arts. The Journal is international in scope and content and publishes papers from a wide range of countries and cultural traditions.