The structures of dynamic and static neural networks for processing stationary and non-stationary data flows during the data mining are examined. These structures make it possible to work with larger time intervals and in a continuous mode, approximate time series and functional dependencies, and perform distributed processing of input data from the electronic warehouse into the neural network.
neural network, structure, continuous mode, distributed processing, electronic warehouse
"Neirosetevoi analyz neprerыvnыkh potokov dannыkh yz эlektronnыkh khranylyshch" ,
Information Processing Systems,