In this study a two phase approach is proposed based on exponential fuzzy time series and learning automata.
In the first phase, the optimal lengths of intervals are estimated by applying LA based EAs in training set.
Second phase aim is to estimate certain adjusting parameters for minimizing errors in training set.
The conventional FTS in the first phase is applied and in the second phase EFTS is employed.
Forty six case studies from five stock index databases are employed in extensive experiments.