|
|
|
|
|
|
|
|
|
Pmse
Глоссарий по искусственному интеллекту |
See: mean square error criterion.
|
|
|
Predictive mean square error, английский
See: mean square error criterion.
Mse criterion, английский
See: mean square error criterion. multi-agent system (mas) see: distributed artificial intelligence. multi-class many classification systems are developed for binary data. multi-class variables are those with more than two categories. this form of variable is also known as multinomial. see also: adaboost.mh, binomial distribution, multi-label. multi-label most class variables belong to a single class. target variables that can belong to more than one class have been called multi-label. an example would be classifications on textual objects that can belong to multiple categories. see also: adaboost.mh, multi-class. multidimensional database (mddb) mddb is a case based reasoning system that is used to diagnosis dysmorphic syndromes, an area with poor medical knowledge by rich information. the system is based on records of over 3000 patients. mddb is commonly used in online analytical processing (olap) and related systems. rather than treat a database of multi-attribute records as a two-dimensional table with rows as records and the attributes (variables) as columns, the data is organized as a k-dimensional rectangle, with one dimension for each attribute. the dimension corresponding to the i -th attribute has ni levels, each corresponding to one of the values that the i -th attribute can assume in this representation. a cell at the intersection of specified values for each of the k attributes contains summary data on all the records that are classified as belonging to that cell. typically, the marginal cells (cells where one or more dimensions have been suppressed) contain information on all of the cells in the missing dimension(s). see also: online analytic processing, http://www.med.unirostock. de/html/english/research/gs52.index.html
|
|
|
|
|
|
|