Export to Machine Learning Services

This page is accessed via Logistics Machine Learning > Machine Learning Project > Actions > Export to Machine Learning Services.

Note: Run the action Load Data into Analytics before running this action to load your shipment data into Logistics Machine Learning Intelligence.

Prerequisites: Several properties must be set for this action to work. They include the URL of the machine learning service and the user name and password for logging into the machine learning service.

Note: This action does not support the use of the Model Type of Order Route Prediction. This will be implemented in a future release.

The Export to Machine Learning action does the following:

1. Extracts your historical shipment data from Logistics Machine Learning intelligence and pre-processes the data into one or more .csv files depending on the Model Type.
   
   If the Model Type is Planned ETA Prediction, then you see a single file containing planned shipment data. If the Model Type is Planned and Event based ETA Prediction, you see two files: one for the planned shipment data and a second for the tracking event data.
2. Zips the file or files into a .gz format.
3. Pushes the file or files to the object store.
4. Notifies the machine learning service that a .gz files are ready for ingestion.
5. Receives responses from the machine learning service and tracks the responses on the Export Data Status page.

You can also run this action via process management.

Exporting Data

Once you select a project and run the action, you see the Export to Machine Learning Service page with important information about the export which includes:

• Project ID
• Number of Records
• Generated Files grid shows:
  • File Names in the format similar to file_otm_eta_shipment_ACB.DOCK-TEST_20201014183606.csv and file_otm_eta_event_ACB.DOCK-TEST_20201014183606.csv
  • Number of Records
  • Size of the .csv file. The maximum size of these files are controlled by the property glog.machinelearning.exportfile.maxrecords.
  • Zip File Names (csv file name above zipped as a .gz file)
  • Size of the zipped files

1. Enter an External System ID that has an Authentication Type of Pre-Authenticated URL. Logistics Machine Learning uses the pre authenticated URL to push the data to an Oracle Cloud Infrastructure or object store.

If you do not specify an external system ID, the export will not work and an error will be logged to the OTM log files.

2. Click Export.
3. Click Refresh to get the latest information from the machine learning service.

When you click Export, the screen automatically updates and may include:

Exported File Information

• File Name of the files exported by Logistics Machine Learning
• Object Store Object Name of the files as stored in object store
• Status indicates if the files were successfully exported to the object store

Data Export Information

• Status indicates if the files were successfully imported by the machine learning service.
• External System ID used to export the file