ML - FloatMLBuilder

ML.FloatMLBuilder

A builder to define (and optionally train) a machine learning model.

Type parameters

Name	Type
`Name`	extends `string`
`Features`	extends `Record`
`Model`	extends `FloatMLModel`

Hierarchy

AbstractMLBuilder

↳ FloatMLBuilder

ML

model

▸ model(model_type):

FloatMLBuilder

Define which MLModel to use.

Type parameters

Name	Type
`M`	extends `FloatMLModel`

Parameters

Name	Type	Description
`model_type`	`M`	the FloatMLModel to apply

Returns

FloatMLBuilder

Example

  // create a gradient boosted tree
  const boosted_tree = new MLModelBuilder("ML Training")
      .feature("x", FloatType)
      .output(FloatType)
      .model({ type: "boosted_tree", noise: "none" })

  // create a gradient boosted tree
  const gaussian_process = new MLModelBuilder("ML Training")
      .feature("x", FloatType)
      .output(FloatType)
      .model({ type: "gaussian_process", noise: "gaussian" })

modelStream

▸ modelStream():

Stream

Return the datastream containing the parameters of trained ML model.

Returns

Stream

The parameter

Stream

Overrides

AbstractMLBuilder.modelStream

toTemplate

▸ toTemplate():

Template

Convert the built ML model into an

Template, for inclusion in an EDK project.

Returns

Template

The

Template containing the ML Model

Overrides

AbstractMLBuilder.toTemplate

trainFromPipeline

▸ trainFromPipeline(config):

FloatMLBuilder

Create a training task based on some input data assembed by a pipeline.

Type parameters

Name	Type
`Output`	extends `string`

Parameters

Name	Type	Description
`config`	`Object`	(optional) the training configuration MLTrainingConfiguration
`config.config?`	`Partial`<`MLTrainingConfiguration`>	-
`config.output_name`	`Output`	-
`config.pipeline`	(`builder`: `PipelineBuilder`) => `TabularPipelineBuilder`<`DictType`<`StringType`, `StructType`<`Features` & { [K in string]: FloatType }>>, `Record`>	-

Returns

FloatMLBuilder

Example

  // use a DictType stream
  const training_stream = Stream(
      "My Stream",
      DictType(
          StringType,
          StructType({
              x: FloatType,
              y: FloatType,
          })
      )
  );

  // create a gradient boosted tree
  const ml_model = new MLModelBuilder("ML Training")
      .feature("x", FloatType)
      .output(FloatType)
      .model({ type: "boosted_trees", noise: "none" })
      .trainFromPipeline({
          output_name: "y",
          pipeline: builder => builder
              .from(training_stream)
              .filter(fields => Greater(fields.x, 0)),
      });

trainFromStream

▸ trainFromStream(config):

FloatMLBuilder

Create a training task based on some input data in an existing datastream.

Type parameters

Name	Type
`Output`	extends `string`

Parameters

Name	Type	Description
`config`	`Object`	the configuration for training (optional)
`config.config?`	`Partial`<`MLTrainingConfiguration`>	-
`config.input`	`Stream`<`DictType`<`StringType`, `StructType`<`Features` & { [K in string]: FloatType }>>>	-
`config.output_name`	`Output`	-

Returns

FloatMLBuilder

Example

  // use a DictType stream
  const training_stream = Stream(
      "My Stream",
      DictType(
          StringType,
          StructType({
              x: FloatType,
              y: FloatType,
          })
      )
  );

  // create a gradient boosted tree
  const ml_model = new MLModelBuilder("ML Training")
      .feature("x", FloatType)
      .output(FloatType)
      .model({ type: "boosted_trees", noise: "none" })
      .trainFromStream({
          output_name: "y",
          input: training_stream,
      });

trainingStream

▸ trainingStream():

Stream<DictType<StringType, StructType<{ features: StructType ; output: FloatType ; test: BooleanType ; train: BooleanType }>>>

Return a datastream containing the training data associated predictions.

Returns

Stream<DictType<StringType, StructType<{ features: StructType ; output: FloatType ; test: BooleanType ; train: BooleanType }>>>

The training

Stream

Overrides

AbstractMLBuilder.trainingStream

validationStream

▸ validationStream():

Stream<DictType<StringType, StructType<{ hyperparameters: DictType ; n_test: IntegerType ; n_train: IntegerType ; test_rmse: FloatType ; test_stddev: FloatType ; train_rmse: FloatType ; train_stddev: FloatType }>>>

Return a datastream containing training statistics per training iteration

Returns

The validation

Stream