PLS Model block
The aim of Partial Least Squares modeling (PLS) is to predict an output data set (Y) from an input data set (X). PLS models are linear models that can predict a specified variable in terms of other observed variables.
PLS regression is used to find the relationship between two data matrices, and the PLS model attempts to find the multidimensional direction in the X space that explains the greatest multidimensional variance direction in the Y space.
Description
The PLS Model block contains one input port and one output port. The input port must contain all fields used in creating the PLS model. These required fields are mapped to the double fields loaded from a source. The output fields will contain the pre-defined target field(s) of the existing PLS model. The PLS model itself is loaded from an XML file exported from the Discrete Troubleshooter where the model was constructed. PLS models can be constructed on all data, inside history brushed data or outside history brushed data. The data can then be normalized and the X Variance contribution set.
PLS MODEL block
Block Type
Rules and Models block
Input port
The input port may contain only double type fields and must contain all fields used in constructing the loaded PLS model. All input fields must be mapped to the required fields of the PLS model.
Output port
The output port will contain a timestamp and the associated prediction of the target field(s) selected when constructing the PLS model.
Functions performed on tags
The input values are sent through the PLS Model block and the calculated output values from the model are placed in output port of the PLS Model block.
-
On the Values: These output values are the predicted target field values, based on the selected model input values.
-
On the Quality: any bad quality input value will result in a bad quality output value.
-
On the Timestamp: It will be recorded at the model execute time.
Runnability
Minimum requirements for runnability on the PLS block:
-
Loaded and valid PLS XML model file.
-
Properly configured and mapped model/input fields. The input port must contain all fields used in constructing the loaded PLS model, and these input fields must be mapped to the required fields of the PLS model.
Related topics: