Prediction of auditor's selectoin by using Heuristic Algoritms in Tehran Stock Exchange

The aim of this study is to predict the independent auditor selected by the Tehran Stock Exchange using the Heuristic Algoritms. The statistical research community includes all companies listed in Tehran Stock Exchange from 2005 until 2010 and the sample is 570 data from 95 corporations for six consecutive years. Data set in the hybrid model algorithm Particle Swarm Optimization with Imperialist Competitive, Artificial Neural networks Multi layered Perceptron and Simulated Annealing Algorithms for prediction of auditor's selection. The results show that all of algorithms determined total assets, current assets, account recivable and work in capital are important in auditor's selection. Other findings show that total debt and audit fee not important for auditor's selection. Other results show that non-audit fee, current rate, debt rate, leverage rate and capital in 2 of 3 algorithms are important in auditor's selection.