public final class IntegratedGradientsAttribution extends GeneratedMessageV3 implements IntegratedGradientsAttributionOrBuilder
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
Protobuf type google.cloud.aiplatform.v1beta1.IntegratedGradientsAttribution
Static Fields
public static final int BLUR_BASELINE_CONFIG_FIELD_NUMBER
| Field Value |
|---|
| Type | Description |
int | |
public static final int SMOOTH_GRAD_CONFIG_FIELD_NUMBER
| Field Value |
|---|
| Type | Description |
int | |
public static final int STEP_COUNT_FIELD_NUMBER
| Field Value |
|---|
| Type | Description |
int | |
Static Methods
public static IntegratedGradientsAttribution getDefaultInstance()
public static final Descriptors.Descriptor getDescriptor()
public static IntegratedGradientsAttribution.Builder newBuilder()
public static IntegratedGradientsAttribution.Builder newBuilder(IntegratedGradientsAttribution prototype)
public static IntegratedGradientsAttribution parseDelimitedFrom(InputStream input)
public static IntegratedGradientsAttribution parseDelimitedFrom(InputStream input, ExtensionRegistryLite extensionRegistry)
public static IntegratedGradientsAttribution parseFrom(byte[] data)
| Parameter |
|---|
| Name | Description |
data | byte[]
|
public static IntegratedGradientsAttribution parseFrom(byte[] data, ExtensionRegistryLite extensionRegistry)
public static IntegratedGradientsAttribution parseFrom(ByteString data)
public static IntegratedGradientsAttribution parseFrom(ByteString data, ExtensionRegistryLite extensionRegistry)
public static IntegratedGradientsAttribution parseFrom(CodedInputStream input)
public static IntegratedGradientsAttribution parseFrom(CodedInputStream input, ExtensionRegistryLite extensionRegistry)
public static IntegratedGradientsAttribution parseFrom(InputStream input)
public static IntegratedGradientsAttribution parseFrom(InputStream input, ExtensionRegistryLite extensionRegistry)
public static IntegratedGradientsAttribution parseFrom(ByteBuffer data)
public static IntegratedGradientsAttribution parseFrom(ByteBuffer data, ExtensionRegistryLite extensionRegistry)
public static Parser<IntegratedGradientsAttribution> parser()
Methods
public boolean equals(Object obj)
| Parameter |
|---|
| Name | Description |
obj | Object
|
Overrides
public BlurBaselineConfig getBlurBaselineConfig()
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
.google.cloud.aiplatform.v1beta1.BlurBaselineConfig blur_baseline_config = 3;
public BlurBaselineConfigOrBuilder getBlurBaselineConfigOrBuilder()
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
.google.cloud.aiplatform.v1beta1.BlurBaselineConfig blur_baseline_config = 3;
public IntegratedGradientsAttribution getDefaultInstanceForType()
public Parser<IntegratedGradientsAttribution> getParserForType()
Overrides
public int getSerializedSize()
| Returns |
|---|
| Type | Description |
int | |
Overrides
public SmoothGradConfig getSmoothGradConfig()
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
.google.cloud.aiplatform.v1beta1.SmoothGradConfig smooth_grad_config = 2;
public SmoothGradConfigOrBuilder getSmoothGradConfigOrBuilder()
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
.google.cloud.aiplatform.v1beta1.SmoothGradConfig smooth_grad_config = 2;
public int getStepCount()
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.
int32 step_count = 1 [(.google.api.field_behavior) = REQUIRED];
| Returns |
|---|
| Type | Description |
int | The stepCount.
|
public final UnknownFieldSet getUnknownFields()
Overrides
public boolean hasBlurBaselineConfig()
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
.google.cloud.aiplatform.v1beta1.BlurBaselineConfig blur_baseline_config = 3;
| Returns |
|---|
| Type | Description |
boolean | Whether the blurBaselineConfig field is set.
|
public boolean hasSmoothGradConfig()
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
.google.cloud.aiplatform.v1beta1.SmoothGradConfig smooth_grad_config = 2;
| Returns |
|---|
| Type | Description |
boolean | Whether the smoothGradConfig field is set.
|
| Returns |
|---|
| Type | Description |
int | |
Overrides
protected GeneratedMessageV3.FieldAccessorTable internalGetFieldAccessorTable()
Overrides
public final boolean isInitialized()
Overrides
public IntegratedGradientsAttribution.Builder newBuilderForType()
protected IntegratedGradientsAttribution.Builder newBuilderForType(GeneratedMessageV3.BuilderParent parent)
Overrides
protected Object newInstance(GeneratedMessageV3.UnusedPrivateParameter unused)
Overrides
public IntegratedGradientsAttribution.Builder toBuilder()
public void writeTo(CodedOutputStream output)
Overrides
