Stay one forget multiple extreme learning machine with deep network using time interval process: A review

摘要

Data streams are the sequence of data packets for communication. The properties of the target variable that is trying to predict, changes at the occurrence of concept drift. So, The observations become less accurate as the time passes. When the speed of concept drift is very fast, In terms of milliseconds, the accuracy of predictions is very difficult to handle. So to solve the problem the new stay one forget multiple extreme learning machine with deep network using time interval process is proposed.