State detecting device, state detecting method, and non-transitory computer-readable medium

A state detecting method, comprising:

• a characteristic quantity detecting process to detect information regarding a monitored subject as characteristic quantities;
• a sorting process to sort multiple classes defining the activity level of a monitored subject into a first group and a second group based on a code table determined by the number of identification devices and the number of classes, each group having at least one or more classes, and associate the sorted multiple classes with the characteristic quantities detected in the characteristic quantity detecting process;
• an identification process to receive the characteristic quantities that are associated in the sorting process as input values, identify whether the state of the monitored subject belongs to the first group or the second group based on the input values, and output information including the identification results and credibility level as output values;
• a loss value calculation process to calculate loss values of the each class based on the output values output in the identification process and the code table;
• and
• a class detection process to detect a class to which the state of the monitored subject belongs among the multiple classes based on the loss values of the each class.

A non-transitory computer-readable medium stored a program that allows a computer to execute the following steps:

• a characteristic quantity detecting step for detecting information regarding a monitored subject as characteristic quantities;
• a sorting step for sorting multiple classes defining the activity level of a monitored subject into a first group and a second group based on a code table determined by the number of identification devices and the number of classes, each group having at least one or more classes, and associating the sorted multiple classes with characteristic quantities detected in the characteristic quantity detecting step;
• an identification step for receiving the characteristic quantities that are associated in the sorting step as input values, identifying whether the state of the monitored subject belongs to the first group or the second group based on the input values, and outputting information including the identification results and credibility level as output values;
• a loss value calculation step calculating loss values of each class based on the output values output in the identification step and the code table;
• and
• a class detection step for detecting a class to which the state of the monitored subject belongs among the multiple classes based on the loss values of the each class.