Local explanation fidelity on Wine (RMSE)
0.003RMSELEMON
Evaluation results
| Method | Links | |
|---|---|---|
| LEMONClassifier=Naive Bayes, Kernel width (sigma)=0.12025.12 | 0.003 | |
| mLIMEClassifier=Naive Bayes, Kernel width (sigma)=0.12025.12 | 0.003 | |
| mLIMEClassifier=Random forest, Kernel width (sigma)=0.12025.12 | 0.005 | |
| LEMONClassifier=Neural network, Kernel width (sigma)=0.12025.12 | 0.007 | |
| mLIMEClassifier=Neural network, Kernel width (sigma)=0.12025.12 | 0.008 | |
| LIMEClassifier=Naive Bayes, Kernel width (sigma)=0.12025.12 | 0.009 | |
| mLIMEClassifier=Random forest, Kernel width (sigma)=0.32025.12 | 0.017 | |
| LEMONClassifier=Random forest, Kernel width (sigma)=0.12025.12 | 0.018 | |
| mLIMEClassifier=Neural network, Kernel width (sigma)=0.32025.12 | 0.019 | |
| mLIMEClassifier=Naive Bayes, Kernel width (sigma)=0.32025.12 | 0.024 | |
| LEMONClassifier=Naive Bayes, Kernel width (sigma)=0.32025.12 | 0.026 | |
| mLIMEClassifier=Random forest, Kernel width (sigma)=0.52025.12 | 0.03 | |
| LEMONClassifier=Neural network, Kernel width (sigma)=0.32025.12 | 0.04 | |
| LIMEClassifier=Random forest, Kernel width (sigma)=0.12025.12 | 0.041 | |
| LIMEClassifier=Naive Bayes, Kernel width (sigma)=0.32025.12 | 0.044 | |
| LEMONClassifier=Random forest, Kernel width (sigma)=0.32025.12 | 0.051 | |
| mLIMEClassifier=Random forest, Kernel width (sigma)=1.02025.12 | 0.051 | |
| mLIMEClassifier=Neural network, Kernel width (sigma)=0.52025.12 | 0.051 | |
| LIMEClassifier=Neural network, Kernel width (sigma)=0.12025.12 | 0.064 | |
| mLIMEClassifier=Naive Bayes, Kernel width (sigma)=0.52025.12 | 0.065 | |
| mLIMEClassifier=Random forest, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.065 | |
| mLIMEClassifier=Random forest, Kernel width (sigma)=4.02025.12 | 0.065 | |
| LEMONClassifier=Naive Bayes, Kernel width (sigma)=0.52025.12 | 0.071 | |
| LEMONClassifier=Random forest, Kernel width (sigma)=0.52025.12 | 0.082 | |
| LIMEClassifier=Neural network, Kernel width (sigma)=0.32025.12 | 0.101 | |
| LIMEClassifier=Naive Bayes, Kernel width (sigma)=0.52025.12 | 0.103 | |
| mLIMEClassifier=Neural network, Kernel width (sigma)=1.02025.12 | 0.106 | |
| LIMEClassifier=Random forest, Kernel width (sigma)=0.32025.12 | 0.118 | |
| LEMONClassifier=Random forest, Kernel width (sigma)=1.02025.12 | 0.12 | |
| LEMONClassifier=Random forest, Kernel width (sigma)=4.02025.12 | 0.12 | |
| LEMONClassifier=Random forest, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.124 | |
| LEMONClassifier=Neural network, Kernel width (sigma)=0.52025.12 | 0.142 | |
| LIMEClassifier=Random forest, Kernel width (sigma)=1.02025.12 | 0.156 | |
| LIMEClassifier=Random forest, Kernel width (sigma)=0.52025.12 | 0.185 | |
| mLIMEClassifier=Naive Bayes, Kernel width (sigma)=1.02025.12 | 0.188 | |
| mLIMEClassifier=Neural network, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.202 | |
| mLIMEClassifier=Neural network, Kernel width (sigma)=4.02025.12 | 0.203 | |
| LEMONClassifier=Neural network, Kernel width (sigma)=1.02025.12 | 0.205 | |
| LIMEClassifier=Neural network, Kernel width (sigma)=0.52025.12 | 0.219 | |
| LEMONClassifier=Naive Bayes, Kernel width (sigma)=1.02025.12 | 0.224 | |
| mLIMEClassifier=Naive Bayes, Kernel width (sigma)=4.02025.12 | 0.226 | |
| mLIMEClassifier=Naive Bayes, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.245 | |
| LIMEClassifier=Naive Bayes, Kernel width (sigma)=1.02025.12 | 0.258 | |
| LIMEClassifier=Neural network, Kernel width (sigma)=1.02025.12 | 0.262 | |
| LEMONClassifier=Neural network, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.273 | |
| LEMONClassifier=Naive Bayes, Kernel width (sigma)=4.02025.12 | 0.282 | |
| LEMONClassifier=Naive Bayes, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.303 | |
| LEMONClassifier=Neural network, Kernel width (sigma)=4.02025.12 | 0.339 | |
| LIMEClassifier=Random forest, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.376 | |
| LIMEClassifier=Neural network, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.499 | |
| LIMEClassifier=Random forest, Kernel width (sigma)=4.02025.12 | 0.545 | |
| LIMEClassifier=Naive Bayes, Kernel width (sigma)=3/4 * sqrt(n)2025.12 | 0.652 | |
| LIMEClassifier=Neural network, Kernel width (sigma)=4.02025.12 | 0.811 | |
| LIMEClassifier=Naive Bayes, Kernel width (sigma)=4.02025.12 | 0.848 |