Improvements to AIML Training Data

NEW JOB TO GENERATE MODELS

We have introduced a new scheduled job that allows you to fully automate your AIML Model Training process. This applies for both the Intelligent Cycle Counting and Order Processing/Cycle/Waiting Time metrics, and Market Basket Analysis. This allows you to seamlessly keep their AIML models up to date.

The new job can be seen in the UI as "Run AIML Training Template."

Setting up this job type is simple and straightforward.
You will need to provide the following information:

• Username: Enter a valid username or login.
• AIML Training Template: Choose the desired template from user-friendly AIML Training Template UI.
• AI Metric: Select the template that best suits your preferences and requirements.

Once this job type is successfully run, our system will process the provided information and generate the model. The output metrics will be conveniently captured in the AIML module UI, with each record identified by a unique composite key, ensuring easy access and organization.

USE DEFAULT TEMPLATE FOR GENERATE PREDICTION RUN

The Generate Prediction Run for AIML Metrics Scheduled Jobs is enhanced so that you can now utilize the Default Template whenever you wish, adding more autonomy to your prediction runs.

The new job parameter, "Use Default Template" allows you to input what you would like to reference when generating a prediction run. If the template_name is populated, then this will be used; however, if the "Use Default Template" is populated, it will be used, and the system will find the default template for that metric.

NOTE: If both are populated, the system will return an error.

NEW OPTION IN GENERATE TRAINING DATA 

We have added a job parameter called "Also Train Default Template," which gives you more control over your generation training data and trains the default template. When creating or editing a scheduling job for AIML metrics, you can now enter the "Also Train Default Template" parameter with yes, no, y, or n.

When this job parameter is enabled, the system automatically includes training data for Order Cycle Time, Processing Time, Waiting Time, and Intelligent Cycle Time in your generation processes. This ensures that your system is trained on a more comprehensive dataset, improving its performance and adaptability.

Steps to Enable

You don't need to do anything to enable this feature.