ImageClassification interface

Package:

@azure/arm-machinelearning

Image Classification. Multi-class image classification is used when an image is classified with only a single label from a set of classes - e.g. each image is classified as either an image of a 'cat' or a 'dog' or a 'duck'.

Extends

AutoMLVertical

Properties

limitSettings	[Required] Limit settings for the AutoML job.
modelSettings	Settings used for training the model.
primaryMetric	Primary metrics for classification tasks.
searchSpace	Search space for sampling different combinations of models and their hyperparameters.
sweepSettings	Model sweeping and hyperparameter sweeping related settings.
taskType	[Required] Task type for AutoMLJob.
validationData	Validation data inputs.
validationDataSize	The fraction of training dataset that needs to be set aside for validation purpose. Values between (0.0 , 1.0) Applied when validation dataset is not provided.

Inherited Properties

logVerbosity	Enum for setting log verbosity.
targetColumnName	Target column name: This is prediction values column. Also known as label column name in context of classification tasks.
trainingData	[Required] Training data input.

Property Details

limitSettings

[Required] Limit settings for the AutoML job.

limitSettings: ImageLimitSettings

Property Value

ImageLimitSettings

modelSettings

Settings used for training the model.

modelSettings?: ImageModelSettingsClassification

Property Value

ImageModelSettingsClassification

primaryMetric

Primary metrics for classification tasks.

primaryMetric?: string

Property Value

string

searchSpace

Search space for sampling different combinations of models and their hyperparameters.

searchSpace?: ImageModelDistributionSettingsClassification[]

Property Value

ImageModelDistributionSettingsClassification[]

sweepSettings

Model sweeping and hyperparameter sweeping related settings.

sweepSettings?: ImageSweepSettings

Property Value

ImageSweepSettings

taskType

[Required] Task type for AutoMLJob.

taskType: "ImageClassification"

Property Value

"ImageClassification"

validationData

Validation data inputs.

validationData?: MLTableJobInput

Property Value

MLTableJobInput

validationDataSize

The fraction of training dataset that needs to be set aside for validation purpose. Values between (0.0 , 1.0) Applied when validation dataset is not provided.

validationDataSize?: number

Property Value

number

Inherited Property Details

logVerbosity

Enum for setting log verbosity.

logVerbosity?: string

Property Value

string

Inherited From AutoMLVertical.logVerbosity

targetColumnName

Target column name: This is prediction values column. Also known as label column name in context of classification tasks.

targetColumnName?: string

Property Value

string

Inherited From AutoMLVertical.targetColumnName

trainingData

[Required] Training data input.

trainingData: MLTableJobInput

Property Value

MLTableJobInput

Inherited From AutoMLVertical.trainingData