Uredi

Dijelite putem

AutoML Text Multi-label Classification

  • Članak

This article describes a component in Azure Machine Learning designer.

Use this component to create a machine learning model that is based on the AutoML Text Multi-label Classification.

Multi-label text classification is for use cases where each example may be assigned more than one label, as opposed to single-label multiclass text classification where every example is labeled with the single most probable class.

How to configure

This component trains an NLP classification model on text data. Text classification is a supervised learning task and requires a labeled dataset that includes a label column with a value for all rows.

This model requires a training and a validation dataset. The datasets must be in ML Table format.

  1. Add the AutoML Text Multi-label Classification component to your pipeline.

  2. Specify the Target Column you want the model to output

  3. Specify the Primary Metric you want AutoML to use to measure your model's success.

  4. (Optional) Select the language your dataset consists of. Visit this link for a full list of supported languages.

  5. (Optional) You are able to configure Hyperparameters. Visit this link for a full list of configurable Hyperparameters

  6. (Optional) Job Sweep settings are configurable. Visit this link to learn more about each configurable parameter.

  7. (Optional) Job Limit settings are configurable. Visit this link to learn more about these settings.

Next steps

See the set of components available to Azure Machine Learning.