Skip to main content

Plan View for ML Upload Tasks

In Plan View, you can create ML Upload tasks to upload new training data to your Machine Learning project. These tasks are specified in the right context panel.

重要

To use this task, you must first create a Machine Learning project. Note that this project can be empty (for example, no data uploaded). If you upload data to a project that already has a trained model, you must rerun the Auto Model stage.

Prepare Prediction Data

  1. In Designer Cloud, create a workflow that outputs your new training data. Note that the training data must be a CSV file with the same column names as the original training data.

  2. In Plans, add a Workflow task to your Plan canvas.

  3. Connect the Workflow task to an ML Upload task.

Configure the ML Upload Task

  1. In the right context panel of the ML Upload task, Select your Machine Learning project.

  2. Under Upstream Task Output > Select Task, select the upstream Workflow task that contains your new training data.

  3. Under Upstream Task Output > Select Output, select the output from the upstream Workflow task that contains the training data.

重要

After you run the Plan, Machine Learning deselects your previous model. You must rerun the Auto Model stage of your Machine Learning project to train the model with the new data.