class Google::Apis::MlV1::GoogleCloudMlV1ExplanationConfig
Message holding configuration options for explaining model predictions. There are three feature attribution methods supported for TensorFlow models: integrated gradients, sampled Shapley, and XRAI. [Learn more about feature attributions.](/ai-platform/prediction/docs/ai-explanations/overview)
Attributes
Attributes credit by computing the Aumann-Shapley value taking advantage of the model's fully differentiable structure. Refer to this paper for more details: arxiv.org/abs/1703.01365 Corresponds to the JSON property `integratedGradientsAttribution` @return [Google::Apis::MlV1::GoogleCloudMlV1IntegratedGradientsAttribution]
An attribution method that approximates Shapley values for features that contribute to the label being predicted. A sampling strategy is used to approximate the value rather than considering all subsets of features. Corresponds to the JSON property `sampledShapleyAttribution` @return [Google::Apis::MlV1::GoogleCloudMlV1SampledShapleyAttribution]
Attributes credit by computing the XRAI taking advantage of the model's fully differentiable structure. Refer to this paper for more details: arxiv. org/abs/1906.02825 Currently only implemented for models with natural image inputs. Corresponds to the JSON property `xraiAttribution` @return [Google::Apis::MlV1::GoogleCloudMlV1XraiAttribution]
Public Class Methods
# File lib/google/apis/ml_v1/classes.rb, line 1114 def initialize(**args) update!(**args) end
Public Instance Methods
Update properties of this object
# File lib/google/apis/ml_v1/classes.rb, line 1119 def update!(**args) @integrated_gradients_attribution = args[:integrated_gradients_attribution] if args.key?(:integrated_gradients_attribution) @sampled_shapley_attribution = args[:sampled_shapley_attribution] if args.key?(:sampled_shapley_attribution) @xrai_attribution = args[:xrai_attribution] if args.key?(:xrai_attribution) end