Explaining Black Boxes With a SMILE: Statistical Model-Agnostic Interpretability With Local Explanations

被引：0

作者：

Aslansefat, Koorosh ^{[1
]}

Hashemian, Mojgan ^{[2
]}

Walker, Martin ^{[1
]}

Akram, Mohammed Naveed ^{[3
]}

Sorokos, Ioannis ^{[3
]}

Papadopoulos, Yiannis ^{[4
]}

机构：

[1] Univ Hull, Comp Sci, Kingston Upon Hull HU6 7RX, England

[2] Direct Line Grp Ltd, Leeds LS1 4AZ, England

[3] Fraunhofer Inst Expt Software Engn, D-67663 Kaiserslautern, Germany

[4] Univ Hull, Dependable Intelligent Syst Res Grp, Kingston Upon Hull HU6 7RX, England

来源：

IEEE SOFTWARE | 2024年 / 41卷 / 01期

关键词：

Closed Box; Perturbation Methods; Predictive Models; Gaussian Distribution; Data Models; Machine Learning; Training; Object Object; Use Of Measures; Statistical Measures; Wide Range Of Domains; Growth In Recent Years; Statistical Distance; Variety Of Supports; Linear Model Object Object; Alternative Models Object Object; Model Coefficients; Maximum Distance; Intersection Over Union Object Object Object Object Object Object; Kernel Function; Input Samples; Light Signal Object Object; Part Of The Image; Adversarial Attacks; Random Perturbations; Perturbation Vector; Human Intuition; Game Theory Object Object; Understanding Of Models;

D O I：

10.1109/MS.2023.3321282

中图分类号：

TP31 [计算机软件];

学科分类号：

081202 ; 0835 ;

摘要：

Explainability is a key aspect of improving trustworthiness. We therefore propose SMILE, a new method that builds on previous approaches by making use of statistical distance measures to improve explainability while remaining applicable to a wide range of input data domains.

引用

页码：87 / 97

页数：11

共 50 条

[21] Local Model-Agnostic Explanations for Black-box Recommender Systems Using Interaction Graphs and Link Prediction Techniques
Caro-Martinez, Marta
Jimenez-Diaz, Guillermo
Recio-Garcia, Juan A.
INTERNATIONAL JOURNAL OF INTERACTIVE MULTIMEDIA AND ARTIFICIAL INTELLIGENCE, 2023, 8 (02): : 202 - 212
[22] Learning Model-Agnostic Counterfactual Explanations for Tabular Data
Pawelczyk, Martin
Broelemann, Klaus
Kasneci, Gjergji
WEB CONFERENCE 2020: PROCEEDINGS OF THE WORLD WIDE WEB CONFERENCE (WWW 2020), 2020, : 3126 - 3132
[23] Model-Agnostic Explanations for Decisions Using Minimal Patterns
Asano, Kohei
Chun, Jinhee
Koike, Atsushi
Tokuyama, Takeshi
ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING - ICANN 2019: THEORETICAL NEURAL COMPUTATION, PT I, 2019, 11727 : 241 - 252
[24] Improving Object Recognition in Crime Scenes via Local Interpretable Model-Agnostic Explanations
Farhood, Helia
Saberi, Morteza
Najafi, Mohammad
2021 IEEE 25TH INTERNATIONAL ENTERPRISE DISTRIBUTED OBJECT COMPUTING CONFERENCE WORKSHOPS (EDOCW 2021), 2021, : 90 - 94
[25] Evaluating Local Interpretable Model-Agnostic Explanations on Clinical Machine Learning Classification Models
Kumarakulasinghe, Nesaretnam Barr
Blomberg, Tobias
Lin, Jintai
Leao, Alexandra Saraiva
Papapetrou, Panagiotis
2020 IEEE 33RD INTERNATIONAL SYMPOSIUM ON COMPUTER-BASED MEDICAL SYSTEMS(CBMS 2020), 2020, : 7 - 12
[26] A Model-Agnostic Approach for Explaining the Predictions on Clustered Data
Zhou, Zihan
Sun, Mingxuan
Chen, Jianhua
2019 19TH IEEE INTERNATIONAL CONFERENCE ON DATA MINING (ICDM 2019), 2019, : 1528 - 1533
[27] Pleural effusion diagnosis using local interpretable model-agnostic explanations and convolutional neural network
Nguyen H.T.
Nguyen C.N.T.
Phan T.M.N.
Dao T.C.
IEIE Transactions on Smart Processing and Computing, 2021, 10 (02): : 101 - 108
[28] Applying local interpretable model-agnostic explanations to identify substructures that are responsible for mutagenicity of chemical compounds
Rosa, Lucca Caiaffa Santos
Pimentel, Andre Silva
MOLECULAR SYSTEMS DESIGN & ENGINEERING, 2024, 9 (09): : 920 - 936
[29] MODEL-AGNOSTIC VISUAL EXPLANATIONS VIA APPROXIMATE BILINEAR MODELS
Joukovsky, Boris
Sammani, Fawaz
Deligiannis, Nikos
2023 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, ICIP, 2023, : 1770 - 1774
[30] Unsupervised Anomaly Detection for Financial Auditing with Model-Agnostic Explanations
Kiefer, Sebastian
Pesch, Gunter
ADVANCES IN ARTIFICIAL INTELLIGENCE, KI 2021, 2021, 12873 : 291 - 308

← 1 2 3 4 5 →