A hybrid least squares support vector

machines and GMDH approach for river

fl ow forecasting-This paper proposes a novel hybrid forecasting model,　which combines the group

method　of　data　handling　(GMDH)　and　the　least　squares　support　vector　machine

(LSSVM), known as GLSSVM. The GMDH is used to determine the useful input vari-

ables for LSSVM model and the LSSVM model which works as time series forecasting. 5

In this study the application of GLSSVM for monthly river fl ow forecasting of Selangor

and Bernam River are investigated.　The results of the proposed GLSSVM approach

are compared with the conventional artifi cial neural network (ANN) models, Autoregres-

sive Integrated Moving Average (ARIMA) model, GMDH and LSSVM models using the

long term observations of monthly river fl ow discharge.　The standard statistical, the 10

root mean square error (RMSE) and coe ffi　cient of correlation (R) are employed to eval-

uate the performance of various models developed.　Experiment result indicates that

the hybrid model was powerful tools to mo