This topic describes how to prepare data by using Integrated Development Environment (IDE) to upload data.
A project is created. For more information, see Create a project.
- Data stored in MaxCompute is used by general-purpose algorithm components.
Note When the data size is smaller than 20 MB, we recommend that you use IDE to upload data. When the data size is larger than 20 MB, we recommend that you use the command-line tool to upload data. For more information, see Run Tunnel commands to upload and download data.
- Structured or unstructured data stored in OSS is used by algorithm components of deep learning.
- Log on to the Machine Learning Platform for AI console.
- In the left-side navigation pane, choose PAI Visualization Modeling page., and navigate to the
- Click Machine Learning.
- Upload data.
- In the left-side navigation pane, click Data Source.
- In the lower-left corner, click Create Table.
- In the Create Table dialog box, enter a name in the Table name field, and specify the Lifecycle (Days).
- In the Schema section, click the icon. Then, enter a name in the Column Name column and specify the table schema in the Type column.
- Click Next.
- Click Select File, and follow the instructions to upload local files.
- Click OK.
- Create an experiment.
- In the left-side navigation pane, click Home.
- In the upper-right corner, click New, and select New Experiment.
- In the New Experiment dialog box, enter a name in the Name field, and click OK.
- Configure the data source.
- In the left-side navigation pane, click Components.
- In the Components list, click Data Source/Target. Then, drag and drop the Read MaxCompute Table component onto the canvas.
- Click the Read MaxCompute Table component on the canvas. On the uicontrol Select Table tab on the right side, enter the name of the created table in the Table Name field.
- Click the Fields Information tab to view the Columns, Type, and Value Range of First 100 of the table.
What to do next
After data preparation is complete, you need to preprocess the data. For more information, see Data preprocessing.