In this submission, I demonstrated the problem of time series prediction using fractional least mean square (FLMS) algorithm.
So far no covergence analysis work is available for FLMS. That's why I didn't mentioned the limits for "etaf". In simple explaination eta and etaf are the mixing weights for the gradient term. Usually, a smaller valu of etaf is chosen compare to the eta.
For the selection of "f", power in the range of 0 and 1 greater than 0 less than 1, is valid.
- correction of weight update rule
-correction of error calculation