A prediction is a BigML resource that stores the value predicted by a supervised learning model for the objective field, i.e., the field you want to predict.
When you make a prediction in BigML, the model returns the predicted value along with a performance measure, expressed either as a confidence for classification problems or as an expected error for regression problems. To trust this prediction, you should have previously evaluated the performance of your supervised model, which once trained, can make predictions for previously unseen instances with an unknown label, as explained in this video:
To learn more about predictions, please:
- Read the corresponding prediction sections of each resource of BigML's Dashboard documentation, or check the API documentation for more details on the arguments and properties of a prediction.
- Read this blog post to learn how to put some confidence in your predictions.
- Check these related questions: How do I know whether my predictions are safe?, Why do I need to evaluate my model?, How can I evaluate my model?