Can Predictor replace Empty values based on another column?

Kristine_E
Kristine_E
in Automation

Hi, I have a large dataset with a "Category" column where about 10% of the records are missing a category. I would like to use Predictor or another tool to fill in the empty categories by looking at a Description column and seeing what Category is assigned to other records with the same or similar description. I created a small test Excel file to see if any of the tools work. I tried using Predictor, but it is not producing the desired result. It seems like maybe I am not using the Predictor tool properly. I am sorry, but the documentation on this is not very thorough. Is there a formula or natural language pattern that I should be typing in?

Comments

  • esiu
    esiu

    Hi Kristine,

    To categorize records with missing categories using the "Classifier" or "Predictor," here are the streamlined steps:

    1. Isolate the rows with missing categories using the "Filter" tool. This filtered dataset should be uploaded as a separate file or exported within the app.
    2. Train your model by selecting the Category column as your target in the "Classifier/Predictor" tool. It will treat other columns as features to make its predictions. Once set up, click "Build" to create your model.
    3. For predictions, use the "Export" tool. In its configuration panel, click "Export now," and in the pop-up, change the input to the file with missing categories. Click "Export now", and Symon.AI will predict the missing categories.
    4. You can download the predicted categories from the Exports page.

Categories