The Detection of the Stock Price Manipulation by Hybrid Genetic Algorithm: Artificial Neural Network Model (ANN-GA) and SQDF Model

The purpose of this research is to detect manipulation of stock prices in Tehran Stock Exchange that it has been done through Hybrid Genetic Algorithm-artificial neural network (ANN-GA) model and the Simplified Quadratic Discriminant Function (SQDF) Model. In this study, the variables of price, trading volume and free float stock to match the results of the model and the actual data of price manipulation is used. In the Hybrid Model of Genetic Algorithm-Artificial Neural Networks (ANN-GA), at first data of 316 stock companies from 2009/03/21 to 2013/03/20 on a daily basis, including 966 days were put into the GA model, then; weight of each variable were derived from GA. Next, using these weights, Perceptron neural network was designed, implemented and its efficiency was approved. Then, SQDF model was designed and implemented and its efficiency was verified. In the end, using MAPE[1], RMSE[2]and R2 error measurement, the results of ANN-GA model were compared with those of SQDF model. The results showed that Hybrid model has much better performance and fewer errors than SQDF model in the detection of stock price manipulation and classifying firms into two groups, manipulate and non-manipulate [1]. Artificial Neural Networks-Genetic Algorithm [2]. Simplified Quadratic Discriminant Function

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.