Basic Guide


What is the Model Generator?

The Model Generator (formerly ML Board) is a MAGELLAN BLOCKS service that anyone can use to generate Machine Learning models, even without programming ability or technical expertise.

Machine Learning with BLOCKS is split into two parts: training and prediction.

  1. During the training step, you use past or known data to train a model.

  2. During the prediction step, you use the trained model to make predictions from new data.

You do the training step with the Model Generator service, and the prediction step with the Model Generator prediction (online) BLOCK or Model Generator prediction (batch) BLOCK from the Flow Designer service.

The training step is simply called a training within the Model Generator. During a training, the Model Generator uses a training set and validation set to train and optimize an accurate model.

The training set and validation set are made by dividing up the training data.

The Model Generator uses Google Cloud Machine Learning Engine to implement its Machine Learning functionality.

Refer to the following pages for more detailed instructions on using the Model Generator: