CreateAutoPredictor - Amazon Forecast

Amazon Forecast is no longer available to new customers. Existing customers of Amazon Forecast can continue to use the service as normal. Learn more"

CreateAutoPredictor

Creates an Amazon Forecast predictor.

Important

Amazon Forecast is no longer available to new customers. Existing customers of Amazon Forecast can continue to use the service as normal. Learn more"

Amazon Forecast creates predictors with AutoPredictor, which involves applying the optimal combination of algorithms to each time series in your datasets. You can use CreateAutoPredictor to create new predictors or upgrade/retrain existing predictors.

Creating new predictors

The following parameters are required when creating a new predictor:

  • PredictorName - A unique name for the predictor.

  • DatasetGroupArn - The ARN of the dataset group used to train the predictor.

  • ForecastFrequency - The granularity of your forecasts (hourly, daily, weekly, etc).

  • ForecastHorizon - The number of time-steps that the model predicts. The forecast horizon is also called the prediction length.

When creating a new predictor, do not specify a value for ReferencePredictorArn.

Upgrading and retraining predictors

The following parameters are required when retraining or upgrading a predictor:

  • PredictorName - A unique name for the predictor.

  • ReferencePredictorArn - The ARN of the predictor to retrain or upgrade.

When upgrading or retraining a predictor, only specify values for the ReferencePredictorArn and PredictorName.

Request Syntax

{ "DataConfig": { "AdditionalDatasets": [ { "Configuration": { "string" : [ "string" ] }, "Name": "string" } ], "AttributeConfigs": [ { "AttributeName": "string", "Transformations": { "string" : "string" } } ], "DatasetGroupArn": "string" }, "EncryptionConfig": { "KMSKeyArn": "string", "RoleArn": "string" }, "ExplainPredictor": boolean, "ForecastDimensions": [ "string" ], "ForecastFrequency": "string", "ForecastHorizon": number, "ForecastTypes": [ "string" ], "MonitorConfig": { "MonitorName": "string" }, "OptimizationMetric": "string", "PredictorName": "string", "ReferencePredictorArn": "string", "Tags": [ { "Key": "string", "Value": "string" } ], "TimeAlignmentBoundary": { "DayOfMonth": number, "DayOfWeek": "string", "Hour": number, "Month": "string" } }

Request Parameters

The request accepts the following data in JSON format.

DataConfig

The data configuration for your dataset group and any additional datasets.

Type: DataConfig object

Required: No

EncryptionConfig

An AWS Key Management Service (KMS) key and an AWS Identity and Access Management (IAM) role that Amazon Forecast can assume to access the key. You can specify this optional object in the CreateDataset and CreatePredictor requests.

Type: EncryptionConfig object

Required: No

ExplainPredictor

Create an Explainability resource for the predictor.

Type: Boolean

Required: No

ForecastDimensions

An array of dimension (field) names that specify how to group the generated forecast.

For example, if you are generating forecasts for item sales across all your stores, and your dataset contains a store_id field, you would specify store_id as a dimension to group sales forecasts for each store.

Type: Array of strings

Array Members: Minimum number of 1 item. Maximum number of 10 items.

Length Constraints: Minimum length of 1. Maximum length of 63.

Pattern: ^[a-zA-Z][a-zA-Z0-9_]*

Required: No

ForecastFrequency

The frequency of predictions in a forecast.

Valid intervals are an integer followed by Y (Year), M (Month), W (Week), D (Day), H (Hour), and min (Minute). For example, "1D" indicates every day and "15min" indicates every 15 minutes. You cannot specify a value that would overlap with the next larger frequency. That means, for example, you cannot specify a frequency of 60 minutes, because that is equivalent to 1 hour. The valid values for each frequency are the following:

  • Minute - 1-59

  • Hour - 1-23

  • Day - 1-6

  • Week - 1-4

  • Month - 1-11

  • Year - 1

Thus, if you want every other week forecasts, specify "2W". Or, if you want quarterly forecasts, you specify "3M".

The frequency must be greater than or equal to the TARGET_TIME_SERIES dataset frequency.

When a RELATED_TIME_SERIES dataset is provided, the frequency must be equal to the RELATED_TIME_SERIES dataset frequency.

Type: String

Length Constraints: Minimum length of 1. Maximum length of 5.

Pattern: ^Y|M|W|D|H|30min|15min|10min|5min|1min$

Required: No

ForecastHorizon

The number of time-steps that the model predicts. The forecast horizon is also called the prediction length.

The maximum forecast horizon is the lesser of 500 time-steps or 1/4 of the TARGET_TIME_SERIES dataset length. If you are retraining an existing AutoPredictor, then the maximum forecast horizon is the lesser of 500 time-steps or 1/3 of the TARGET_TIME_SERIES dataset length.

If you are upgrading to an AutoPredictor or retraining an existing AutoPredictor, you cannot update the forecast horizon parameter. You can meet this requirement by providing longer time-series in the dataset.

Type: Integer

Required: No

ForecastTypes

The forecast types used to train a predictor. You can specify up to five forecast types. Forecast types can be quantiles from 0.01 to 0.99, by increments of 0.01 or higher. You can also specify the mean forecast with mean.

Type: Array of strings

Array Members: Minimum number of 1 item. Maximum number of 20 items.

Length Constraints: Minimum length of 2. Maximum length of 4.

Pattern: (^0?\.\d\d?$|^mean$)

Required: No

MonitorConfig

The configuration details for predictor monitoring. Provide a name for the monitor resource to enable predictor monitoring.

Predictor monitoring allows you to see how your predictor's performance changes over time. For more information, see Predictor Monitoring.

Type: MonitorConfig object

Required: No

OptimizationMetric

The accuracy metric used to optimize the predictor.

Type: String

Valid Values: WAPE | RMSE | AverageWeightedQuantileLoss | MASE | MAPE

Required: No

PredictorName

A unique name for the predictor

Type: String

Length Constraints: Minimum length of 1. Maximum length of 63.

Pattern: ^[a-zA-Z][a-zA-Z0-9_]*

Required: Yes

ReferencePredictorArn

The ARN of the predictor to retrain or upgrade. This parameter is only used when retraining or upgrading a predictor. When creating a new predictor, do not specify a value for this parameter.

When upgrading or retraining a predictor, only specify values for the ReferencePredictorArn and PredictorName. The value for PredictorName must be a unique predictor name.

Type: String

Length Constraints: Maximum length of 256.

Pattern: arn:([a-z\d-]+):forecast:.*:.*:.+

Required: No

Tags

Optional metadata to help you categorize and organize your predictors. Each tag consists of a key and an optional value, both of which you define. Tag keys and values are case sensitive.

The following restrictions apply to tags:

  • For each resource, each tag key must be unique and each tag key must have one value.

  • Maximum number of tags per resource: 50.

  • Maximum key length: 128 Unicode characters in UTF-8.

  • Maximum value length: 256 Unicode characters in UTF-8.

  • Accepted characters: all letters and numbers, spaces representable in UTF-8, and + - = . _ : / @. If your tagging schema is used across other services and resources, the character restrictions of those services also apply.

  • Key prefixes cannot include any upper or lowercase combination of aws: or AWS:. Values can have this prefix. If a tag value has aws as its prefix but the key does not, Forecast considers it to be a user tag and will count against the limit of 50 tags. Tags with only the key prefix of aws do not count against your tags per resource limit. You cannot edit or delete tag keys with this prefix.

Type: Array of Tag objects

Array Members: Minimum number of 0 items. Maximum number of 200 items.

Required: No

TimeAlignmentBoundary

The time boundary Forecast uses to align and aggregate any data that doesn't align with your forecast frequency. Provide the unit of time and the time boundary as a key value pair. For more information on specifying a time boundary, see Specifying a Time Boundary. If you don't provide a time boundary, Forecast uses a set of Default Time Boundaries.

Type: TimeAlignmentBoundary object

Required: No

Response Syntax

{ "PredictorArn": "string" }

Response Elements

If the action is successful, the service sends back an HTTP 200 response.

The following data is returned in JSON format by the service.

PredictorArn

The Amazon Resource Name (ARN) of the predictor.

Type: String

Length Constraints: Maximum length of 256.

Pattern: arn:([a-z\d-]+):forecast:.*:.*:.+

Errors

InvalidInputException

We can't process the request because it includes an invalid value or a value that exceeds the valid range.

HTTP Status Code: 400

LimitExceededException

The limit on the number of resources per account has been exceeded.

HTTP Status Code: 400

ResourceAlreadyExistsException

There is already a resource with this name. Try again with a different name.

HTTP Status Code: 400

ResourceInUseException

The specified resource is in use.

HTTP Status Code: 400

ResourceNotFoundException

We can't find a resource with that Amazon Resource Name (ARN). Check the ARN and try again.

HTTP Status Code: 400

See Also

For more information about using this API in one of the language-specific AWS SDKs, see the following: