AVEVA™ Unified Engineering

Machine Learning

Save PDF

Save PDF

Use the functionality to utilize the benefits of a machine learning model in any PML development.

Model Name	STRING	Enter a name for the ML.Net Model (PMLNETCALLABLE object)
Seed	REAL	Enter a seed value (seed value for Microsoft MLContext's random number generator). Note: If a fixed seed is provided, the ML.NET model environment becomes deterministic, meaning that the results are repeatable and will remain the same across multiple runs. Please refer to the Microsoft ML.NET documentation for further information.
ML.NET Location	File Path (STRING)	Enter or browse to the location of the trained ML.Net Model.
PML Callable Location	File Path (STRING)	Enter or browse to the output location for the PMLNETCALLABLE dll.

Note:
The customization of windows and/or functions require administrative permissions.

Methods available on the PMLNETCALLABLE Object

GetSchema()	ARRAY	Returns the input schema for the Microsoft ML.NET model.
MakePrediction(DYNAMIC)	DYNAMIC	Returns the result of the model and the probabilities as an array, based on the parameters passed as input to the model.

Create a PMLNETCALLABLE dll using the Machine Learning window.

Note:
The dll can be used in the same way as any NET customization.

-- Import the dll

import 'C:\DLLS\MLNET'

using namespace 'MLNET'

-- Load object

!mlModel = OBJECT MLNET()

-- Make prediction on the ML.NET model. This will have arguments

-- based on the ML.NET model trained.

!prediction = !mlModel.makePrediction(7.2, 3.6, 6.1, 2.5)