From e9e4b8850dde2005d408ed2f003b8f6e32529397 Mon Sep 17 00:00:00 2001 From: Sebastian Fischer Date: Wed, 16 Oct 2024 17:18:11 +0200 Subject: [PATCH] ... --- book/chapters/chapter12/model_interpretation.qmd | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/book/chapters/chapter12/model_interpretation.qmd b/book/chapters/chapter12/model_interpretation.qmd index fabcc17ff..e6f343d9c 100644 --- a/book/chapters/chapter12/model_interpretation.qmd +++ b/book/chapters/chapter12/model_interpretation.qmd @@ -315,7 +315,7 @@ The `sample.size` argument (default is `sample.size = 100`) can be increased to #| fig-cap: Shapley values for Charlie. The actual prediction (0.63) displays the prediction of the model for the observation we are interested in, the average prediction (0.71) displays the average prediction over the given test dataset. Each horizontal bar is the Shapley value (phi) for the given feature. #| fig-alt: 10 bar plots of Shapley values, one for each feature. x-axis says 'phi' and ranges from -0.1 to 0.05. The strongest positive contributions are from the `duration`, `purpose` and `property` variables. The strongest negative contributions are `status`, `amount`, and `savings`. #| label: fig-iml-shapley -shapley = Shapley$new(predictor, x.interest = Charlie, +shapley = Shapley$new(predictor, x.interest = as.data.frame(Charlie), sample.size = 1000) shapley$plot() ```