Selection of inference models based on geographic information
An information processing device 1 X mainly includes a selection means 32 X and an inference means 35 X. The selection means 32 X is configured to select, based on one or more geographical center points according to a movement history of an inference target device, one or more adaptive inference models to be used from inference models generated on an area-by-area basis, the inference models making inferences of a state of the inference target device. The inference means 35 X is configured to make an inference of the state of the inference target device based on inference results of the adaptive inference models which use data acquired by the inference target device.
1 . An information processing device comprising:
at least one memory configured to store instructions; and
at least one processor configured to execute the instructions to:
calculate geographical center points for different time periods of movement history of a vehicle, respectively;
select, a plurality of adaptive inference models for each of the geographical center points, respectively, from inference models generated on an area-by-area basis, the inference models inferring a state of the vehicle;
obtain intermediate inference results by integrating inference results for each of the geographical center points based on the selected adaptive inference models;
obtain a final inference result by weighting the intermediate inference results of each of the geographical center points based on length of history and integrating the weighted intermediate results for each of the geographical center points; and
control the vehicle to drive based on the final inference result.
2 . The information processing device according to claim 1 ,
wherein the time period is determined within a time period after a previous maintenance for the target inference device.
3 . The information processing device according to claim 1 ,
wherein the inference models generated on the area-by-area basis are learning models learned on an area-by-area basis based on training data acquired for each area.
4 . The information processing device according to claim 1 ,
wherein the at least one processor is configured to execute the instructions to perform at least one of abnormality detection of the state, classification of the state, or calculation of a score representing the state as the inference of the state.
5 . The information processing device according to claim 1 ,
wherein the at least one processor is configured to further execute the instructions, upon detecting an abnormality of the state of the vehicle, to
output information on movement of the vehicle causing the abnormality or
perform a control on the vehicle.
6 . The information processing device according to claim 5 ,
wherein the at least one processor is configured to execute the instructions to highlight a travelling section causing the abnormality on a movement locus of the vehicle based on the movement history as an output of the information on the movement of the vehicle.
7 . The information processing device according to claim 5 ,
wherein the at least one processor is configured to execute the instructions to perform restriction on the movement of the inference vehicle support as the control on the vehicle.
8 . The information processing device according to claim 1 ,
wherein the at least one processor is configured to execute the instructions to determine a receiver of the inference result on the state in accordance with the inference result on the state.
9 . The information processing device according to claim 1 ,
wherein the at least one processor is configured to execute the instructions to infer the state of the vehicle in whole or one or more components of the vehicle.
10 . The information processing device according to claim 1 ,
wherein the at least one processor is configured to execute the instructions to select the adaptive inference models based on
a distance between each of the center points and one or more areas in which a data center in change of generating inference models is located, or
a degree of similarity of environmental conditions between each of the center points and the one or more areas.
11 . The information processing device according to claim 1 ,
wherein the at least one processor is configured to execute the instructions to acquire, as the data, at least one of
output data from a sensor provided in the vehicle or
control data generated in the vehicle.
12 . The information processing device according to claim 1 ,
wherein the information processing device is
a part of the vehicle, or
one or more external devices configured to make the inference of the state of the vehicle by receiving the data from the vehicle.
13 . A control method executed by a computer, the control method comprising:
calculating geographical center points for different time periods of movement history of a vehicle, respectively;
selecting, a plurality of adaptive inference models for each of the geographical center points, respectively, from inference models generated on an area-by-area basis, the inference models inferring a state of the vehicle;
obtaining intermediate inference results by integrating inference results for each of the geographical center points based on the selected adaptive inference models;
obtaining a final inference result by weighting the intermediate inference results of each of the geographical center points based on length of history and integrating the weighted intermediate results for each of the geographical center points; and
controlling the vehicle to drive based on the final inference result.
14 . A non-transitory computer readable storage medium storing a program executed by a computer, the program causing the computer to perform processing comprising:
calculating geographical center points for different time periods of movement history of a vehicle, respectively;
selecting, a plurality of adaptive inference models for each of the geographical center points, respectively, from inference models generated on an area-by-area basis, the inference models inferring a state of the vehicle;
obtaining intermediate inference results by integrating inference results for each of the geographical center points based on the selected adaptive inference models;
obtaining a final inference result by weighting the intermediate inference results of each of the geographical center points based on length of history and integrating the weighted intermediate results for each of the geographical center points; and
controlling the vehicle to drive based on the final inference result.
15 . The information processing device according to claim 3 , wherein the learning models are generated by machine learning based on the training data.
16 . The information processing device according to claim 1 , wherein the processor is further configured to: generate, based on the inference of the state, information for supporting decision making of a user or a manager regarding maintenance of the vehicle; and output the generated information.