External stimuli predict financial market behavior from the brain perception perspective

Abstract

The information that the brain perceives is usually consistent with a range of possible incentives. Therefore, all of our perceptual decisions are almost made in an uncertain situation. As we all know, this uncertainty affects our behavior, but how this uncertainty to modify human behavior is unclear. We attempt to establish the relationship between financial market behavior and external stimulus information. We adopt a new approach that is entirely different from the existing literature. This approach combines neuroscience and machine learning methods to explore how the brain perceives external stimulus information and ultimately influences financial market behavior. We improve the BP neural network in two aspects. Firstly, the output of the brain perception model serves as the input of the BP neural network. By this method, the number of input nodes of the BP neural network can be reduced to six, and the mental process behind the stimulus is simulated. Secondly, we optimize the parameters of the brain perception model and construct the optimal brain perception model for specific external stimuli. By comparing the performance of all models, the results show that the improved BP neural network is superior to other models. Firstly, in all two periods, trends are similar between the improved BP neural network and other models. Secondly, in all three samples, except for one result, the average prediction performance of the improved BP neural network is better than other models.