Enhanced Financial Asset Price Prediction Using Multi-Layer Perceptron: A Deep Learning Approach for Modeling Nonlinear Market Dynamics

Abstract

This paper proposes a financial asset price prediction model based on multi-layer perceptron (MLP), which models financial market data through deep learning methods. Financial asset price prediction has always been an important topic in the financial field. Although traditional prediction methods such as linear regression and support vector machine have achieved certain results in some scenarios, they often cannot effectively capture the complex nonlinear relationships in financial data. To make up for this deficiency, this paper adopts the MLP model, which can automatically learn the deep features and laws in the data through a multi-layer network structure and nonlinear activation function. Experimental results show that the MLP-based model outperforms traditional methods in multiple evaluation indicators, especially in terms of mean square error (MSE), root mean square error (RMSE) and mean absolute error (MAE). This study not only verifies the effectiveness of MLP in financial asset price prediction, but also demonstrates the great potential of deep learning in processing complex financial data. Future research can further optimize the model on this basis, explore the combination of multimodal data and reinforcement learning and other technologies, further improve the prediction accuracy, and promote the development of intelligent decision-making in the financial field.