Enhancing the explainability of regression-based polynomial chaos expansion by Shapley additive explanations

Palar, Pramudita Satria; Zuhal, Lavi Rizki; Shimoyama, Koji

doi:10.1016/j.ress.2022.109045

Cited by 25 publications

(6 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…SHAP is based on an approach subject to cooperative game theory [ 20 ], which centers on the introduction of Shapley values to construct an additive explanatory model that facilitates the interpretation of the performance of machine learning models or deep learning models, i.e., Shapley values [ 21 ] measure the marginal contribution of each input variable in the overall cooperation. This approach falls under the category of additive feature attribution, and a surprising property in this category is the existence of a single unique solution that has three desirable properties: local accuracy, missingness, and consistency.…”

Section: Methodsmentioning

confidence: 99%

Machine Learning Models Using SHapley Additive exPlanation for Fire Risk Assessment Mode and Effects Analysis of Stadiums

Fan

Zhang

et al. 2023

Sensors

View full text Add to dashboard Cite

Machine learning methods can establish complex nonlinear relationships between input and response variables for stadium fire risk assessment. However, the output of machine learning models is considered very difficult due to their complex “black box” structure, which hinders their application in stadium fire risk assessment. The SHapley Additive exPlanations (SHAP) method makes a local approximation to the predictions of any regression or classification model so as to be faithful and interpretable, and assigns significant values (SHAP value) to each input variable for a given prediction. In this study, we designed an indicator attribute threshold interval to classify and quantify different fire risk category data, and then used a random forest model combined with SHAP strategy in order to establish a stadium fire risk assessment model. The main objective is to analyze the impact analysis of each risk characteristic on four different risk assessment models, so as to find the complex nonlinear relationship between risk characteristics and stadium fire risk. This helps managers to be able to make appropriate fire safety management and smart decisions before an incident occurs and in a targeted manner to reduce the incidence of fires. The experimental results show that the established interpretable random forest model provides 83% accuracy, 86% precision, and 85% recall for the stadium fire risk test dataset. The study also shows that the low level of data makes it difficult to identify the range of decision boundaries for Critical mode and Hazardous mode.

show abstract

Section: Methodsmentioning

confidence: 99%

Machine Learning Models Using SHapley Additive exPlanation for Fire Risk Assessment Mode and Effects Analysis of Stadiums

Fan

Zhang

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…The SHAP value is used as a unified measure in measuring feature importance. This Shapley value is the value of the conditional expectation function of the original model [100], [101]. Thus, they are solutions to the equation:…”

Section: Shap Values (Shapley Additive Explanations)mentioning

confidence: 99%

Boosting Algorithm to Handle Unbalanced Classification of PM_2.5 Concentration Levels by Observing Meteorological Parameters in Jakarta-Indonesia Using AdaBoost, XGBoost, CatBoost, and LightGBM

et al. 2023

View full text Add to dashboard Cite

show abstract

“…The first aerodynamic problem is the GSA of a subsonic wing to investigate the effect of wing twists on aerodynamic performance [42]. The input variables are the twist angle at eight different sections of an untapered wing with NACA 2412 airfoil, varied from -10 to 10 degrees.…”

Section: B Test Case 1: 8-variable Subsonic Wing Analysis With Vortex...mentioning

confidence: 99%

Global Sensitivity Analysis in Aerodynamic Design Using Shapley Effects and Polynomial Chaos Regression

Palar,

Zuhal,

Shimoyama

2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

Quantifying the impact of design variables in aerodynamic design exploration can provide valuable insights to designers. Global sensitivity analysis (GSA) is a crucial tool in aerodynamic design exploration that enables designers to gain valuable insights by quantifying the impact of design variables. In the field of GSA, the Shapley effect is a powerful alternative to total Sobol indices due to several mathematical advantages of the former. However, computing the Shapley effect is computationally expensive due to the large number of permutations involved. To overcome this challenge, surrogate models are often used to accurately estimate Shapley effects while reducing the number of function calls. This paper aims to investigate the effectiveness of using PCE to compute Shapley effects for independent inputs in aerodynamic design exploration. The exact calculation from PCE also enables the rapid assessment of confidence intervals for Shapley effects, taking into account the randomness in the experimental design via bootstrap resampling. The usefulness of Shapley effects with PCE is then demonstrated and compared with total Sobol indices through a nonlinear test function and three engineering problems, including subsonic wing, transonic airfoil, and fan blade design. The results also show that the confidence intervals of the Shapley effects are narrower than those of total Sobol indices, allowing better interpretation and higher confidence on the estimated GSA metric.

show abstract

Enhancing the explainability of regression-based polynomial chaos expansion by Shapley additive explanations

Cited by 25 publications

References 36 publications

Machine Learning Models Using SHapley Additive exPlanation for Fire Risk Assessment Mode and Effects Analysis of Stadiums

Machine Learning Models Using SHapley Additive exPlanation for Fire Risk Assessment Mode and Effects Analysis of Stadiums

Boosting Algorithm to Handle Unbalanced Classification of PM_2.5 Concentration Levels by Observing Meteorological Parameters in Jakarta-Indonesia Using AdaBoost, XGBoost, CatBoost, and LightGBM

Global Sensitivity Analysis in Aerodynamic Design Using Shapley Effects and Polynomial Chaos Regression

Contact Info

Product

Resources

About

Enhancing the explainability of regression-based polynomial chaos expansion by Shapley additive explanations

Cited by 25 publications

References 36 publications

Machine Learning Models Using SHapley Additive exPlanation for Fire Risk Assessment Mode and Effects Analysis of Stadiums

Machine Learning Models Using SHapley Additive exPlanation for Fire Risk Assessment Mode and Effects Analysis of Stadiums

Boosting Algorithm to Handle Unbalanced Classification of PM2.5 Concentration Levels by Observing Meteorological Parameters in Jakarta-Indonesia Using AdaBoost, XGBoost, CatBoost, and LightGBM

Global Sensitivity Analysis in Aerodynamic Design Using Shapley Effects and Polynomial Chaos Regression

Contact Info

Product

Resources

About

Boosting Algorithm to Handle Unbalanced Classification of PM_2.5 Concentration Levels by Observing Meteorological Parameters in Jakarta-Indonesia Using AdaBoost, XGBoost, CatBoost, and LightGBM