Options

How should the results of the complex process be linked to the AI hub's dashboard?

kimjkkimjk Member Posts: 19 Maven
edited April 2023 in Help
https://academy.rapidminer.com/learn/video/dashboards-01-data-and-webservice

As shown in the video above, the process of simply importing data from the dashboard has been successful.
However, it is not possible to load the dataset in which the label was predicted by calculating in the process and training the model.
Please let me know if I need to add some operators or if there is another way.  thank you

Best Answer

  • Options
    vivek101vivek101 Member Posts: 8 Contributor II
    Solution Accepted
    Hi @kimjk

    You can use the procedures provided in the video tutorial you cited to connect the outcomes of a complicated procedure to the AI hub's dashboard in RapidMiner. But here are some actions you may do if you're having trouble loading the dataset with projected labels:

    1. Verify the format and structure of the dataset containing the expected labels. Verify that the projected labels column has the exact same name and data type as the original dataset.

    2. Verify that the predicted labels, if you're using them, have been correctly connected to the original dataset, if you're using a different data source. The "Join" operator can be used to merge the datasets based on a shared key.

    3. Verify that the label prediction model is correctly trained and preserved. Retrain the model or save it once more if necessary.

    4. Verify that the appropriate parameters, such as the data source and file location, have been set if you're using the "Retrieve" operator to import the dataset.

    5. If none of the aforementioned procedures are successful, think about using a different approach to load the projected labels, such as a database or web service.

    After loading the dataset with the predicted labels successfully, you can use the same technique as in the video instruction to build a dashboard and view the outcomes of your intricate procedure.

    Thank you.
Sign In or Register to comment.