| CPC A63B 24/0062 (2013.01) [A63B 24/0003 (2013.01); G06N 7/00 (2013.01)] | 11 Claims |
|
1. An information processing system, comprising:
a wearable sensor device configured to be worn on a body of a subject and which includes:
a motion sensor configured to detect motion of the subject,
a position sensor configured to acquire a current position of the subject based on a GPS signal, and
a condition determination sensor for detecting condition information related to a condition under which the subject performs the motion; and
at least one processor configured to:
generate time-series exercise data based on output from the wearable sensor device, wherein, in generating the time-series exercise data, the at least one processor calculates a plurality of indicators based on at least acceleration data output from the motion sensor and position data output from the position sensor, the plurality of indicators including at least a first indicator expressing a first exercise metric and a second indicator which has a correlative relationship with the first indicator and which expresses a second exercise metric other than the first exercise metric;
generate a first period-specific model using the time-series exercise data from a first time period which is selected based on the condition information, wherein the first period-specific model is trained to express a relationship between the first indicator and the second indicator in the first time period using values of the first indicator in the time-series exercise data from the first time period as input and using values of the second indicator in the time-series exercise data from the first time period as output, and is structured to receive a value of the first indicator as an input and to generate a value of the second indicator as an output;
generate a second period-specific model using the time-series exercise data from a second time period other than the first time period and which is selected based on the condition information, wherein the second period-specific model is trained to express a relationship between the first indicator and the second indicator in the second time period using values of the first indicator in the time-series exercise data from the second time period as input and using values of the second indicator in the time-series exercise data from the second time period as output, and is structured to receive a value of the first indicator as an input and to generate a value of the second indicator as an output;
store the generated first period-specific model and second period-specific model in a memory;
receive, via a user interface, input of a value of the first indicator;
generate first estimation data by inputting the value of the first indicator received via the user interface into the stored first period-specific model, to obtain an estimated value of the second indicator that corresponds to the input value of the first indicator in the first time period as the first estimation data;
generate second estimation data by inputting the value of the first indicator received via the user interface into the stored second period-specific model, to obtain an estimated value of the second indicator that corresponds to the input value of the first indicator in the second time period as the second estimation data;
perform display control to display, on a display, the generated first estimation data and the generated second estimation data in a mutually comparable graphical format.
|