How to Open a Black Box Classifier for Tabular Data
被引:6
|
作者:
Walters, Bradley
论文数: 0引用数: 0
h-index: 0
机构:
Liverpool John Moores Univ, Sch Comp Sci & Math, Liverpool L3 2AF, EnglandLiverpool John Moores Univ, Sch Comp Sci & Math, Liverpool L3 2AF, England
Walters, Bradley
[1
]
论文数: 引用数:
h-index:
机构:
Ortega-Martorell, Sandra
[1
]
Olier, Ivan
论文数: 0引用数: 0
h-index: 0
机构:
Liverpool John Moores Univ, Sch Comp Sci & Math, Liverpool L3 2AF, EnglandLiverpool John Moores Univ, Sch Comp Sci & Math, Liverpool L3 2AF, England
Olier, Ivan
[1
]
Lisboa, Paulo J. G.
论文数: 0引用数: 0
h-index: 0
机构:
Liverpool John Moores Univ, Sch Comp Sci & Math, Liverpool L3 2AF, EnglandLiverpool John Moores Univ, Sch Comp Sci & Math, Liverpool L3 2AF, England
Lisboa, Paulo J. G.
[1
]
机构:
[1] Liverpool John Moores Univ, Sch Comp Sci & Math, Liverpool L3 2AF, England
A lack of transparency in machine learning models can limit their application. We show that analysis of variance (ANOVA) methods extract interpretable predictive models from them. This is possible because ANOVA decompositions represent multivariate functions as sums of functions of fewer variables. Retaining the terms in the ANOVA summation involving functions of only one or two variables provides an efficient method to open black box classifiers. The proposed method builds generalised additive models (GAMs) by application of L1 regularised logistic regression to the component terms retained from the ANOVA decomposition of the logit function. The resulting GAMs are derived using two alternative measures, Dirac and Lebesgue. Both measures produce functions that are smooth and consistent. The term partial responses in structured models (PRiSM) describes the family of models that are derived from black box classifiers by application of ANOVA decompositions. We demonstrate their interpretability and performance for the multilayer perceptron, support vector machines and gradient-boosting machines applied to synthetic data and several real-world data sets, namely Pima Diabetes, German Credit Card, and Statlog Shuttle from the UCI repository. The GAMs are shown to be compliant with the basic principles of a formal framework for interpretability.
机构:
Charite Univ Med Berlin, Ctr Hlth Data Sci, Berlin Inst Hlth, Med Informat Grp, Berlin, GermanyCharite Univ Med Berlin, Ctr Hlth Data Sci, Berlin Inst Hlth, Med Informat Grp, Berlin, Germany
Haber, Anna C.
Sax, Ulrich
论文数: 0引用数: 0
h-index: 0
机构:
Univ Med Ctr Gottingen, Dept Med Informat, Med Informat, Gottingen, Germany
Univ Med Ctr Gottingen, Dept Med Informat, Res Grp Infrastruct Translat Res, Gottingen, GermanyCharite Univ Med Berlin, Ctr Hlth Data Sci, Berlin Inst Hlth, Med Informat Grp, Berlin, Germany
Sax, Ulrich
Prasser, Fabian
论文数: 0引用数: 0
h-index: 0
机构:
Charite Univ Med Berlin, Ctr Hlth Data Sci, Berlin Inst Hlth, Med Informat Grp, Berlin, GermanyCharite Univ Med Berlin, Ctr Hlth Data Sci, Berlin Inst Hlth, Med Informat Grp, Berlin, Germany