Developer Docs: What happens during Setup?

Initiating the cloud_classifier class

During setup, a cloud_classifier class is initiated via:

cc = cloud_classifier.cloud_classifier()

This triggers two actions:

1. the cloud_project class is initiated

2. the cloud_trainer class is initiated

Initiating the cloud_project class

The init of the cloud project handles:

1. setting the project_path

2. initiating a parameter_handler class

3. retrieves the default parameters from the parameter_handler

4. initializes an index set for masking

Initiating the parameter_handler class

This sets default parameters:

• it defines json config files for general setting aspects

• it defines filelists again as json files, e.g. input files, label files, etc.

• it defines parameters for the machine learning method + details on the input and label characteristics

Initiating the cloud_trainer class

The cloud_trainer class is responsible for

1. training an ML model based on input and label data

2. applying a trained ML model to new input data

During initialization no further action is taken.