Machine learning method and machine learning apparatus performing learning relating to work process

A machine learning apparatus for performing learning relating to a plurality of different work processes respectively including at least single types of unit work operations, the machine learning apparatus comprising:

• a work totaling part configured to total up all of the unit work operations included in the plurality of different work processes so as to assess a presence of a plurality of unit work operations of the same type of the unit work operations;
• a similarity judging part configured to judge whether the plurality of unit work operations of the same type are similar to each other, wherein the similarity judging part is configured to
• calculate a number of work elements common among the plurality of unit work operations of the same type among the work elements of the plurality of unit work operations,
• calculate a degree of similarity from the number of work elements common among the plurality of unit work operations of the same type, and
• determine whether the degree of similarity is a predetermined value or more;
• a first learning part configured to
• define a first unit work operation and a second unit work operation that are similar to each other as a set of similar work operations, and
• use a common machine learning algorithm so as to generate a similar work learning model using features of the set of similar work operations as input data;
• and
• a second learning part configured to perform learning relating to a first work process including the first unit work operation and a second work process including the second unit work operation based on the similar work learning model, wherein
• the first learning part is configured to
• acquire, as a learning data set, status data of data including content information of the unit work operations and showing the features and label data showing whether actions of the unit work operations corresponding to the status data are abnormal,
• calculate a threshold value for judging whether the actions of the unit work operations are susceptible to abnormality based on the learning data set,
• in response to an action of the actions of the unit work operations being susceptible to abnormality, cause a display to display normal work information before the action is performed,
• in response to the action being abnormal after causing the display to display the normal work information, instruct, through the display, to correct the action being abnormal, and
• generate the similar work learning model showing correlation between the content information and the threshold value using the learning data set.

A machine learning apparatus for performing learning relating to a plurality of different work processes respectively including at least single types of unit work operations, the machine learning apparatus comprising:

• a processor configured to
• total up all of the unit work operations included in the plurality of different work processes so as to assess a presence of a plurality of unit work operations of the same type of the unit work operations;
• judge whether the plurality of unit work operations of the same type are similar to each other, wherein the judge includes
• calculating a number of work elements common among the plurality of unit work operations of the same type among the work elements of the plurality of unit work operations,
• calculating a degree of similarity from the number of work elements common among the plurality of unit work operations of the same type, and
• determining whether the degree of similarity is a predetermined value or more;
• define a first unit work operation and a second unit work operation that are similar to each other as a set of similar work operations;
• use a common machine learning algorithm so as to generate a similar work learning model using features of the set of similar work operations as input data;
• and
• perform learning relating to a first work process including the first unit work operation and a second work process including the second unit work operation based on the similar work learning model, wherein
• the processor is configured to
• acquire, as a learning data set, status data of data including content information of the unit work operations and showing the features and label data showing whether actions of the unit work operations corresponding to the status data are abnormal,
• calculate a threshold value for judging whether the actions of the unit work operations are susceptible to abnormality based on the learning data set,
• in response to an action of the actions of the unit work operations being susceptible to abnormality, cause a display to display normal work information before the action is performed,
• in response to the action being abnormal after causing the display to display the normal work information, instruct, through the display, to correct the action being abnormal, and
• generate the similar work learning model showing correlation between the content information and the threshold value using the learning data set.