Pages: 109-117

Abstract

Predicting stock price has special importance for shareholders to gain the maximum profit and they have always sought for logical and accurate strategies to predict it. Data mining techniques, in addition to data collection and management, involve analyzing and predicting. Recognizing the current patterns and unknown relationships among the data help us in the predicting. Several models have been developed for predicting by using time series by researchers in the recent years. Given the studies conducted in this regard, it can be realized that one of the important issues in these models is the way of determining the fuzzy intervals to explain the model and to predict. Three models were introduced in this research using combination of fuzzy time series and cuckoo optimization algorithms (FTS-COA), and combination of fuzzy time series and particle swarm algorithm (FTS-PSO), and combination of fuzzy time series and firefly algorithm (FTS-FOA). Finally, to compare the introduced models, findings of these three models are compared. Findings reflect the superiority of the (FTS-COA) model compared to the two models introduced.