Regression models are used to predict continuous values. They provide a function that determines a relationship between an independent variable and a dependent variable, allowing you to find trends in your data.
Plainsight supports labeling images using a regression label type that supports a numerical value. In the future, an API will be available to supply data for training a regression model.
Labeled image for a regression model
Define a Regression Label
1.
Navigate to the "Label Definitions" tab of your dataset
2.
Type label name in the "Name" field
3.
Choose Regression from "Type" dropdown
4.
Click "Save" button
Data Format
A Regression schema element creates an array of objects with data consisting of a float value.
1
{
2
"type": "regression",
3
"data": 168,
4
"children": {}
5
}
Copied!
Labeling Regression Data
Once you've defined your Regression label, you can begin annotating your regression data.
You can train a regression model by specifying Regression as the Model Output Type in your SmartML configuration. Check out our walk-through on building your regression model in Plainsight: