Implementation of Online sequential extreme learning machine with kernels for nonstationary time series prediction.
In this work, an OS-ELMK was studied and coded in Octave. To illustrate results, simulations were done with artificial and real nonstationary time series. Also, the ELMK was implemented for comparative analysis.
The original article - https://www.sciencedirect.com/science/article/abs/pii/S0925231214007711