جستجوی مقالات مرتبط با کلیدواژه « multilayer perceptron neural network » در نشریات گروه « مالی »

The purpose of this study is predict and design Credit risk model for debut crowdfunding .According, the complexity of the risk assessment the best neural network architecture with Customize hidden layer neurons selected Multilayer perceptron algorithm for simulation. The statistical population of this study is the financial information of credit / loan file of all customer (506 cases) one of the banks of the country for the year 1997-98. In order to show the significant relationship the extracted indices of the sample and the model output variables (non-default and default), the sample member tested by regression.Thus, thirteen indices entered to the model neural network input vector with three hidden layers in non-default and default groups. In the simulation results, the proposed model was able to optimize the weights of each of the inputs to the network with lower prediction error and 94.1% efficiency .also the average error absolute value obtained for training data (0.88), test data (0.94) and evaluation data (0.84) indicating high capability of the proposed model. According to the research Results, among the indices, income, 0.163 weight, Current Account weight 0.123 are more important, but “degree of education of education” 0.053 are less important in the non-defaulted group.

Stock trading systems development using evolutionary algorithms over the past few years has become a hot topic in financial fields. In this paper, an intelligent technical trading system was proposed using a combination of MLP neural network and evolutionary algorithms (i.e., GA, ACOR, and PSO). In order to select the final variables as the selected features, a return comparison of each indicator ratings was used based on tradings. Finally, the performance of each model is tested in comparison with the buy and hold strategy. The results show that the evolutionary learning algorithms significantly outperform the benchmark models in terms of the average return and the hybrid MLP_PSO model outperforms others.

The issue of behavioral finance is one of the new debates raised by some financial pundits over the past two decades. The unknown factors affecting stock price changes are always a reason to use stock price prediction. In most predictive models, the system performs prediction using only one indicator, but in the proposed model in this study, a two-level system of multilayered perceptron neural networks is presented, which uses several indicators for prediction. In this study, required information of Tehran stock exchange price indicators, for fiscal years 2012 - 2017 was collected. In order to analyze the financial behavior, after examining the effect of each behavioral factor on the investment of financial assets, the results show that all factors other than "over-confidence" affect investment, but the effectiveness of each factor, including "relative profit and loss", " disposition effect", "conservatism", "herd behavior", " representativeness", " ownership" and " regret aversion" is different. Among these factors, the "relative profit and loss" has had the most impact on the investment of financial assets in the stock exchange, and the " regret aversion" has the least, which proceeding a direct impact on the stock price index. Also, for better training of the neural network and consequently improving the results, grasshopper optimization algorithm is used to select the best samples. The results show that the proposed model could have lower predictive error than other models.

Due to the complexity of the stock market and the high volume of processable information, often using a simple system to predict cannot release appropriate results. Therefore, researchers have been trying to provide a system with less complexity and more efficiency and accuracy using hybrid models. nowadays various patters are used including statistical technique (discriminate analysis , logistic , analysis factors) and artificial intelligent techniques ( neural networks(NN) , decision trees , case based reasoning , genetic algorithm , rough sets , support vector machine , fuzzy logic ) and the combination of these two technique for predicating stock prices. For most predictive models, the system uses only one indicator to predict, but in the proposed model in this study, a two-level system of multilayered perceptron neural networks is presented which uses several indicators to predict. To do this, required information of Tehran Stock Exchange price indicators, for fiscal years 2012 - 2017 was collected. We also used the Grasshopper Optimization Algorithm to select the best samples for better nerve network training and thus to improve the results.  The results show that the proposed model can operate with lower prediction error than other models.