Class GoogleCloudAiplatformV1IntegratedGradientsAttribution
An attribution method that computes the Aumann-Shapley value taking advantage of the model's fully differentiable structure. Refer to this paper for more details: https://arxiv.org/abs/1703.01365
Implements
Inherited Members
Namespace: Google.Apis.Aiplatform.v1.Data
Assembly: Google.Apis.Aiplatform.v1.dll
Syntax
public class GoogleCloudAiplatformV1IntegratedGradientsAttribution : IDirectResponseSchema
Properties
BlurBaselineConfig
Config for IG with blur baseline. When enabled, a linear path from the maximally blurred image to the input image is created. Using a blurred baseline instead of zero (black image) is motivated by the BlurIG approach explained here: https://arxiv.org/abs/2004.03383
Declaration
[JsonProperty("blurBaselineConfig")]
public virtual GoogleCloudAiplatformV1BlurBaselineConfig BlurBaselineConfig { get; set; }
Property Value
Type | Description |
---|---|
GoogleCloudAiplatformV1BlurBaselineConfig |
ETag
The ETag of the item.
Declaration
public virtual string ETag { get; set; }
Property Value
Type | Description |
---|---|
string |
SmoothGradConfig
Config for SmoothGrad approximation of gradients. When enabled, the gradients are approximated by averaging the gradients from noisy samples in the vicinity of the inputs. Adding noise can help improve the computed gradients. Refer to this paper for more details: https://arxiv.org/pdf/1706.03825.pdf
Declaration
[JsonProperty("smoothGradConfig")]
public virtual GoogleCloudAiplatformV1SmoothGradConfig SmoothGradConfig { get; set; }
Property Value
Type | Description |
---|---|
GoogleCloudAiplatformV1SmoothGradConfig |
StepCount
Required. The number of steps for approximating the path integral. A good value to start is 50 and gradually increase until the sum to diff property is within the desired error range. Valid range of its value is [1, 100], inclusively.
Declaration
[JsonProperty("stepCount")]
public virtual int? StepCount { get; set; }
Property Value
Type | Description |
---|---|
int? |