Systems and methods for generating a structural model architecture
OwnerMarcus Kaiser, Rui Sampaio +2inventors
InventorsMarcus Kaiser, Rui Sampaio, Andrew Lawrence, Maksim Sipos
Life of this patent
Estimating the term…
Abstract
A method of generating a structural model architecture, comprising receiving an input causality map, including: a plurality of variables, one or more links providing an indication of influence between pairs of variables, wherein at least one link provides an indication of influence between an input variable and at least one other variable; and one or more constraints associated with one or more of the variables and/or links, and generating an architecture for a structural model based on the input causality map, the structural model configured to predict a value of an output variable based on a value of at least one other variable in the plurality of variables.
Claims15
A computer-implemented method of generating a structural model architecture, comprising one or more processors, wherein the method further comprises:
- receiving, by the processor, an input causality map from training data, including:
- a plurality of variables;
- one or more links providing an indication of influence between pairs of variables, wherein at least one link provides an indication of influence between an input variable and at least one other variable;
- and
- one or more constraints associated with one or more of the variables and/or links, wherein the one or more constraints include a limit, a symmetry and a property of the associated variables or links;
- and
- generating, by the processor an architecture for a structural model based on the input causality map, the structural model configured to predict a value of an output variable based on a value of at least one other variable in the plurality of variables.
Dependencies are inferred from the claim text.