{"title":"试图用机器学习超越因果关系:探索性研究中模型可解释性技术的局限性。","authors":"Matthew J Vowels","doi":"10.1037/met0000699","DOIUrl":null,"url":null,"abstract":"Machine learning explainability techniques have been proposed as a means for psychologists to \"explain\" or interrogate a model in order to gain an understanding of a phenomenon of interest. Researchers concerned with imposing overly restrictive functional form (e.g., as would be the case in a linear regression) may be motivated to use machine learning algorithms in conjunction with explainability techniques, as part of exploratory research, with the goal of identifying important variables that are associated with/predictive of an outcome of interest. However, and as we demonstrate, machine learning algorithms are highly sensitive to the underlying causal structure in the data. The consequences of this are that predictors which are deemed by the explainability technique to be unrelated/unimportant/unpredictive, may actually be highly associated with the outcome. Rather than this being a limitation of explainability techniques per se, we show that it is rather a consequence of the mathematical implications of regression, and the interaction of these implications with the associated conditional independencies of the underlying causal structure. We provide some alternative recommendations for psychologists wanting to explore the data for important variables. (PsycInfo Database Record (c) 2024 APA, all rights reserved).","PeriodicalId":20782,"journal":{"name":"Psychological methods","volume":"9 1","pages":""},"PeriodicalIF":7.6000,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Trying to outrun causality with machine learning: Limitations of model explainability techniques for exploratory research.\",\"authors\":\"Matthew J Vowels\",\"doi\":\"10.1037/met0000699\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Machine learning explainability techniques have been proposed as a means for psychologists to \\\"explain\\\" or interrogate a model in order to gain an understanding of a phenomenon of interest. Researchers concerned with imposing overly restrictive functional form (e.g., as would be the case in a linear regression) may be motivated to use machine learning algorithms in conjunction with explainability techniques, as part of exploratory research, with the goal of identifying important variables that are associated with/predictive of an outcome of interest. However, and as we demonstrate, machine learning algorithms are highly sensitive to the underlying causal structure in the data. The consequences of this are that predictors which are deemed by the explainability technique to be unrelated/unimportant/unpredictive, may actually be highly associated with the outcome. Rather than this being a limitation of explainability techniques per se, we show that it is rather a consequence of the mathematical implications of regression, and the interaction of these implications with the associated conditional independencies of the underlying causal structure. We provide some alternative recommendations for psychologists wanting to explore the data for important variables. (PsycInfo Database Record (c) 2024 APA, all rights reserved).\",\"PeriodicalId\":20782,\"journal\":{\"name\":\"Psychological methods\",\"volume\":\"9 1\",\"pages\":\"\"},\"PeriodicalIF\":7.6000,\"publicationDate\":\"2024-09-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Psychological methods\",\"FirstCategoryId\":\"102\",\"ListUrlMain\":\"https://doi.org/10.1037/met0000699\",\"RegionNum\":1,\"RegionCategory\":\"心理学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"PSYCHOLOGY, MULTIDISCIPLINARY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Psychological methods","FirstCategoryId":"102","ListUrlMain":"https://doi.org/10.1037/met0000699","RegionNum":1,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"PSYCHOLOGY, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
摘要
机器学习可解释性技术已被提出作为心理学家 "解释 "或询问模型的一种手段,以获得对相关现象的理解。研究人员可能会担心强加过于严格的函数形式(如线性回归中的函数形式),因此会将机器学习算法与可解释性技术结合起来使用,作为探索性研究的一部分,目的是找出与感兴趣的结果相关/可预测结果的重要变量。然而,正如我们所展示的,机器学习算法对数据中的潜在因果结构非常敏感。其结果是,可解释性技术认为不相关/不重要/不可预测的预测因子,实际上可能与结果高度相关。与其说这是可解释性技术本身的局限性,不如说是回归的数学含义以及这些含义与基本因果结构的相关条件独立性相互作用的结果。我们为希望探索重要变量数据的心理学家提供了一些替代建议。(PsycInfo Database Record (c) 2024 APA, 版权所有)。
Trying to outrun causality with machine learning: Limitations of model explainability techniques for exploratory research.
Machine learning explainability techniques have been proposed as a means for psychologists to "explain" or interrogate a model in order to gain an understanding of a phenomenon of interest. Researchers concerned with imposing overly restrictive functional form (e.g., as would be the case in a linear regression) may be motivated to use machine learning algorithms in conjunction with explainability techniques, as part of exploratory research, with the goal of identifying important variables that are associated with/predictive of an outcome of interest. However, and as we demonstrate, machine learning algorithms are highly sensitive to the underlying causal structure in the data. The consequences of this are that predictors which are deemed by the explainability technique to be unrelated/unimportant/unpredictive, may actually be highly associated with the outcome. Rather than this being a limitation of explainability techniques per se, we show that it is rather a consequence of the mathematical implications of regression, and the interaction of these implications with the associated conditional independencies of the underlying causal structure. We provide some alternative recommendations for psychologists wanting to explore the data for important variables. (PsycInfo Database Record (c) 2024 APA, all rights reserved).
期刊介绍:
Psychological Methods is devoted to the development and dissemination of methods for collecting, analyzing, understanding, and interpreting psychological data. Its purpose is the dissemination of innovations in research design, measurement, methodology, and quantitative and qualitative analysis to the psychological community; its further purpose is to promote effective communication about related substantive and methodological issues. The audience is expected to be diverse and to include those who develop new procedures, those who are responsible for undergraduate and graduate training in design, measurement, and statistics, as well as those who employ those procedures in research.