Graph-based predictive maintenance

1. A method for detecting failure components, comprising:

performing, by a processor device, graph-based predictive maintenance (GBPM) to determine a trained ensemble classification model for detecting maintenance ready components that includes extracted node features and graph features;

constructing, based on testing data and the trained ensemble classification model, an attributed temporal graph and the extracted node features and graph features;

concatenating the extracted node features and graph features; and

determining, based on the trained ensemble classification model, a list of prediction results of components that are to be scheduled for component maintenance;

wherein the graph features include temporal graph features; and

wherein the temporal graph features are selected from the group consisting of: in-degree from the attributed the temporal graph; total in-degree from the attributed temporal graph and at least one unknown component; out-degree from the attributed temporal graph; and total out-degree from attributed temporal graph and the at least one unknown component.

2. A method for detecting failure components, comprising:

performing, by a processor device, graph-based predictive maintenance (GBPM) to determine a trained ensemble classification model for detecting maintenance ready components that includes extracted node features and graph features;

constructing, based on testing data and the trained ensemble classification model, an attributed temporal graph and the extracted node features and graph features;

concatenating the extracted node features and graph features; and

determining, based on the trained ensemble classification model, a list of prediction results of components that are to be scheduled for component maintenance;

wherein the graph features include temporal graph features; and

wherein the temporal graph features include a dark web out-degree from unknown companies to existing component.

3. A method for detecting failure components, comprising:

performing, by a processor device, graph-based predictive maintenance (GBPM) to determine a trained ensemble classification model for detecting maintenance ready components that includes extracted node features and graph features;

constructing, based on testing data and the trained ensemble classification model, an attributed temporal graph and the extracted node features and graph features;

concatenating the extracted node features and graph features; and

determining, based on the trained ensemble classification model, a list of prediction results of components that are to be scheduled for component maintenance;

wherein the graph features include temporal graph features; and

wherein the temporal graph features are selected from the group consisting of:

difference between total in-degree and total out-degree; and

proportion between total in-degree and total out-degree.

4. A computer system for predicting system device failure, comprising:

a processor device operatively coupled to a memory device, the processor device being configured to:

perform graph-based predictive maintenance (GBPM) to determine a trained ensemble classification model for detecting maintenance ready components that includes extracted node features and graph features;

construct, based on testing data and the trained ensemble classification model, an attributed temporal graph and the extracted node features and graph features;

concatenate the extracted node features and graph features; and

determine, based on the trained ensemble classification model, a list of prediction results of components that are to be scheduled for component maintenance;

wherein the graph features include temporal graph features; and

wherein the temporal graph features are selected from the group consisting of:

in-degree from the attributed the temporal graph;

total in-degree from the attributed temporal graph and at least one unknown component;

out-degree from the attributed temporal graph; and

total out-degree from attributed temporal graph and the at least one unknown component.

5. A computer system for predicting system device failure, comprising:

a processor device operatively coupled to a memory device, the processor device being configured to:

perform graph-based predictive maintenance (GBPM) to determine a trained ensemble classification model for detecting maintenance ready components that includes extracted node features and graph features;

construct, based on testing data and the trained ensemble classification model, an attributed temporal graph and the extracted node features and graph features;

concatenate the extracted node features and graph features; and

determine, based on the trained ensemble classification model, a list of prediction results of components that are to be scheduled for component maintenance;

wherein the graph features include temporal graph features; and

wherein the temporal graph features include a dark web out-degree from unknown companies to existing component.

6. A computer system for predicting system device failure, comprising:

a processor device operatively coupled to a memory device, the processor device being configured to:

perform graph-based predictive maintenance (GBPM) to determine a trained ensemble classification model for detecting maintenance ready components that includes extracted node features and graph features;

construct, based on testing data and the trained ensemble classification model, an attributed temporal graph and the extracted node features and graph features;

concatenate the extracted node features and graph features; and

determine, based on the trained ensemble classification model, a list of prediction results of components that are to be scheduled for component maintenance;

wherein the temporal graph features are selected from the group consisting of:

difference between total in-degree and total out-degree; and proportion between total in-degree and total out-degree.