Statistical Hypothesis Testing versus Machine Learning Binary Classification: Distinctions and Guidelines.
Statistical Hypothesis Testing versus Machine Learning Binary Classification: Distinctions and Guidelines.
复制标题
DOI：
10.1016/j.patter.2020.100115
复制
发表时间：
2020-10-09
期刊：
影响因子：
--
通讯作者：
Tong X
中科院分区：
文献类型：
--
作者：
Li JJ;Tong X
进行二元决策是科学研究和工业应用中常见的数据分析任务。在数据科学中，有两种相关但不同的策略：假设检验和二元分类。在实践中，如何在这两种策略之间做出选择可能不清楚，而且相当令人困惑。在这里，我们总结了这两种策略在三个方面的主要区别，并列出了数据分析师为特定分析需求选择适当策略的五个实用指南。我们在癌症驱动基因预测的例子中展示了这些指导方针的使用。在数据科学教育中，假设检验和二元分类这两种分析策略主要在不同的课程和教科书中涉及。在真实的数据应用中，二元决策问题究竟应表述为假设检验还是二元分类是一个令人困惑的问题。本文旨在为数据科学专业的学生和研究人员解开这个难题，为在这两种策略之间进行选择提供实用的指导。假设检验和二元分类是两种数据分析策略，主要在不同的本科课程中教授，很少相互比较。因此，哪种策略更适合于特定的现实世界数据分析任务通常是模糊的。为了解决这个问题，本文阐明了这两种策略之间的区别，并为广泛的数据科学学科提供了实用的指导方针。
Making binary decisions is a common data analytical task in scientific research and industrial applications. In data sciences, there are two related but distinct strategies: hypothesis testing and binary classification. In practice, how to choose between these two strategies can be unclear and rather confusing. Here, we summarize key distinctions between these two strategies in three aspects and list five practical guidelines for data analysts to choose the appropriate strategy for specific analysis needs. We demonstrate the use of those guidelines in a cancer driver gene prediction example. In data science education, two analysis strategies, hypothesis testing and binary classification, are mostly covered in different courses and textbooks. In real data application, it can be puzzling whether a binary decision problem should be formulated as hypothesis testing or binary classification. This article aims to disentangle the puzzle for data science students and researchers by offering practical guidelines for choosing between the two strategies. Hypothesis testing and binary classification are two data analysis strategies taught mostly in different undergraduate classes and rarely compared with each other. As a result, which strategy is more appropriate for a specific real-world data analysis task is often ambiguous. To address this issue, this perspective article clarifies the distinctions between the two strategies and offers practical guidelines to the broad data science discipline.
登录
查看更多内容
影响因子：
64.5
作者：
Davoli T;Xu AW;Mengwasser KE;Sack LM;Yoon JC;Park PJ;Elledge SJ
通讯作者：
Elledge SJ
DOI：
10.1111/j.2517-6161.1995.tb02031.x
发表时间：
1995-01-01
影响因子：
5.8
作者：
BENJAMINI, Y;HOCHBERG, Y
通讯作者：
HOCHBERG, Y
影响因子：
16.6
作者：
Branco, Paula;Torgo, Luis;Ribeiro, Rita P.
通讯作者：
Ribeiro, Rita P.
影响因子：
5.7
作者：
Breiman, L
通讯作者：
Breiman, L
影响因子：
46.9
作者：
Bansal, Mukesh;Yang, Jichen;Karan, Charles;Menden, Michael P.;Costello, James C.;Tang, Hao;Xiao, Guanghua;Li, Yajuan;Allen, Jeffrey;Zhong, Rui;Chen, Beibei;Kim, Minsoo;Wang, Tao;Heiser, Laura M.;Realubit, Ronald;Mattioli, Michela;Alvarez, Mariano J.;Shen, Yao;Gallahan, Daniel;Singer, Dinah;Saez-Rodriguez, Julio;Xie, Yang;Stolovitzky, Gustavo;Califano, Andrea;Abbuehl, Jean-Paul;Altman, Russ B.;Balcome, Shawn;Bell, Ana;Bender, Andreas;Berger, Bonnie;Bernard, Jonathan;Bieberich, Andrew A.;Borboudakis, Giorgos;Chan, Christina;Chen, Ting-Huei;Choi, Jaejoon;Coelho, Luis Pedro;Creighton, Chad J.;Dampier, Will;Davisson, V. Jo;Deshpande, Raamesh;Diao, Lixia;Di Camillo, Barbara;Dundar, Murat;Ertel, Adam;Goswami, Chirayu P.;Gottlieb, Assaf;Gould, Michael N.;Goya, Jonathan;Grau, Michael;Gray, Joe W.;Hejase, Hussein A.;Hoffmann, Michael F.;Homicsko, Krisztian;Homilius, Max;Hwang, Woochang;Ijzerman, Adriaan P.;Kallioniemi, Olli;Karacali, Bilge;Kaski, Samuel;Kim, Junho;Krishnan, Arjun;Lee, Junehawk;Lee, Young-Suk;Lenselink, Eelke B.;Lenz, Peter;Li, Lang;Li, Jun;Liang, Han;Mpindi, John-Patrick;Myers, Chad L.;Newton, Michael A.;Overington, John P.;Parkkinen, Juuso;Prill, Robert J.;Peng, Jian;Pestell, Richard;Qiu, Peng;Rajwa, Bartek;Sadanandam, Anguraj;Sambo, Francesco;Sridhar, Arvind;Sun, Wei;Toffolo, Gianna M.;Tozeren, Aydin;Troyanskaya, Olga G.;Tsamardinos, Ioannis;van Vlijmen, Herman W. T.;Wang, Wen;Wegner, Joerg K.;Wennerberg, Krister;van Westen, Gerard J. P.;Xia, Tian;Yang, Yang;Yao, Victoria;Yuan, Yuan;Zeng, Haoyang;Zhang, Shihua;Zhao, Junfei;Zhou, Jian
通讯作者：
Zhou, Jian