publicfinalclass IntegratedGradientsAttributionextendsGeneratedMessageV3implementsIntegratedGradientsAttributionOrBuilder
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.vertexai.v1.IntegratedGradientsAttribution
Inherited Members
com.google.protobuf.GeneratedMessageV3.<ListT>makeMutableCopy(ListT)
com.google.protobuf.GeneratedMessageV3.<ListT>makeMutableCopy(ListT,int)
com.google.protobuf.GeneratedMessageV3.<T>emptyList(java.lang.Class<T>)
com.google.protobuf.GeneratedMessageV3.internalGetMapFieldReflection(int)
Static Fields
BLUR_BASELINE_CONFIG_FIELD_NUMBER
publicstaticfinalintBLUR_BASELINE_CONFIG_FIELD_NUMBER
| Field Value |
| Type |
Description |
int |
SMOOTH_GRAD_CONFIG_FIELD_NUMBER
publicstaticfinalintSMOOTH_GRAD_CONFIG_FIELD_NUMBER
| Field Value |
| Type |
Description |
int |
STEP_COUNT_FIELD_NUMBER
publicstaticfinalintSTEP_COUNT_FIELD_NUMBER
| Field Value |
| Type |
Description |
int |
Static Methods
getDefaultInstance()
publicstaticIntegratedGradientsAttributiongetDefaultInstance()
getDescriptor()
publicstaticfinalDescriptors.DescriptorgetDescriptor()
newBuilder()
publicstaticIntegratedGradientsAttribution.BuildernewBuilder()
newBuilder(IntegratedGradientsAttribution prototype)
publicstaticIntegratedGradientsAttribution.BuildernewBuilder(IntegratedGradientsAttributionprototype)
publicstaticIntegratedGradientsAttributionparseDelimitedFrom(InputStreaminput)
publicstaticIntegratedGradientsAttributionparseDelimitedFrom(InputStreaminput,ExtensionRegistryLiteextensionRegistry)
parseFrom(byte[] data)
publicstaticIntegratedGradientsAttributionparseFrom(byte[]data)
| Parameter |
| Name |
Description |
data |
byte[]
|
parseFrom(byte[] data, ExtensionRegistryLite extensionRegistry)
publicstaticIntegratedGradientsAttributionparseFrom(byte[]data,ExtensionRegistryLiteextensionRegistry)
parseFrom(ByteString data)
publicstaticIntegratedGradientsAttributionparseFrom(ByteStringdata)
parseFrom(ByteString data, ExtensionRegistryLite extensionRegistry)
publicstaticIntegratedGradientsAttributionparseFrom(ByteStringdata,ExtensionRegistryLiteextensionRegistry)
publicstaticIntegratedGradientsAttributionparseFrom(CodedInputStreaminput)
publicstaticIntegratedGradientsAttributionparseFrom(CodedInputStreaminput,ExtensionRegistryLiteextensionRegistry)
publicstaticIntegratedGradientsAttributionparseFrom(InputStreaminput)
publicstaticIntegratedGradientsAttributionparseFrom(InputStreaminput,ExtensionRegistryLiteextensionRegistry)
parseFrom(ByteBuffer data)
publicstaticIntegratedGradientsAttributionparseFrom(ByteBufferdata)
parseFrom(ByteBuffer data, ExtensionRegistryLite extensionRegistry)
publicstaticIntegratedGradientsAttributionparseFrom(ByteBufferdata,ExtensionRegistryLiteextensionRegistry)
parser()
publicstaticParser<IntegratedGradientsAttribution>parser()
Methods
equals(Object obj)
publicbooleanequals(Objectobj)
| Parameter |
| Name |
Description |
obj |
Object
|
Overrides
getBlurBaselineConfig()
publicBlurBaselineConfiggetBlurBaselineConfig()
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.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
getBlurBaselineConfigOrBuilder()
publicBlurBaselineConfigOrBuildergetBlurBaselineConfigOrBuilder()
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.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
getDefaultInstanceForType()
publicIntegratedGradientsAttributiongetDefaultInstanceForType()
getParserForType()
publicParser<IntegratedGradientsAttribution>getParserForType()
Overrides
getSerializedSize()
publicintgetSerializedSize()
| Returns |
| Type |
Description |
int |
Overrides
getSmoothGradConfig()
publicSmoothGradConfiggetSmoothGradConfig()
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.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
getSmoothGradConfigOrBuilder()
publicSmoothGradConfigOrBuildergetSmoothGradConfigOrBuilder()
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.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
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.
|
hasBlurBaselineConfig()
publicbooleanhasBlurBaselineConfig()
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.vertexai.v1.BlurBaselineConfig blur_baseline_config = 3;
| Returns |
| Type |
Description |
boolean |
Whether the blurBaselineConfig field is set.
|
hasSmoothGradConfig()
publicbooleanhasSmoothGradConfig()
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.vertexai.v1.SmoothGradConfig smooth_grad_config = 2;
| Returns |
| Type |
Description |
boolean |
Whether the smoothGradConfig field is set.
|
hashCode()
| Returns |
| Type |
Description |
int |
Overrides
internalGetFieldAccessorTable()
protectedGeneratedMessageV3.FieldAccessorTableinternalGetFieldAccessorTable()
Overrides
isInitialized()
publicfinalbooleanisInitialized()
Overrides
newBuilderForType()
publicIntegratedGradientsAttribution.BuildernewBuilderForType()
newBuilderForType(GeneratedMessageV3.BuilderParent parent)
protectedIntegratedGradientsAttribution.BuildernewBuilderForType(GeneratedMessageV3.BuilderParentparent)
Overrides
newInstance(GeneratedMessageV3.UnusedPrivateParameter unused)
protectedObjectnewInstance(GeneratedMessageV3.UnusedPrivateParameterunused)
| Returns |
| Type |
Description |
Object |
Overrides
toBuilder()
publicIntegratedGradientsAttribution.BuildertoBuilder()
writeTo(CodedOutputStream output)
publicvoidwriteTo(CodedOutputStreamoutput)
Overrides