青云翻译

Learning problems in a multi-tag, because each object may also have multiple category tags, so the traditional supervised learning commonly used single marker evaluation indicators, such as precision (accuracy), precision rate (precision), the recall rate (recall) and so on, can not be directly used

Tag learning problems because each object may have multiple category tags at, so traditional tag commonly used in supervised learning evaluation indicators, such as precision (accuracy) and precision (precision), searching rate (recall), and so cannot be used directly in tag performance evaluation o

In multi-mark study question, because each object possibly simultaneously has many category mark, therefore the traditional surveillance studies the commonly used list mark appraisal target, like precision (accuracy), accuracy ratio (precision), recall (recall) and so on, is unable to use in marking

Learn more than the mark in question, as each object may also have more than one category, and the traditional tags commonly used in the study of single-tag evaluation indicators, such as precision (accuracy), the Associate rate (precision), verification, such as the full rate (recall) cannot be dir

Tag learning problems because each object may have multiple category tags at, so traditional tag commonly used in supervised learning evaluation indicators, such as precision (accuracy) and precision (precision), searching rate (recall), and so cannot be used directly in tag performance evaluation o