Combining Deep Learning and Multiresolution Analysis for Stock Market Forecasting

Abstract

Due to its complexity, financial time-series forecasting is regarded as one of the most challenging problems. During the past two decades, nonlinear modeling techniques, such as artificial neural networks, were commonly employed to solve a variety of time-series problems. Recently, however, deep neural network has been found to be more efficient than those in many application domains. In this article, we propose a model based on deep neural networks that improves the forecasting of stock prices. We investigate the impact of combining deep learning techniques with multiresolution analysis to improve the forecasting accuracy. Our proposed model is based on an empirical wavelet transform which is shown to outperform traditional stationary wavelet transform in capturing price fluctuations at different time scales. The proposed model is demonstrated to be substantially more effective than other models when evaluated on the S&P500 stock index and Mackey-Glass time series.

Modeling current and future role of agricultural waste in the production of bioethanol for gasoline vehicles

“…In addition, unexpected and external events in the economic, social, industrial, political, and environmental sectors cause a change in the predicted trends in the fuel demand model. The use of complex nonlinear modeling systems such as artificial neural networks to predict chaotic time series models such as meteorology (Altan et al 2021), medical sciences (Ramirez-Carrasco and Molina-Garay 2021), and stock markets (Althelaya et al 2021) have a long history. The main purpose of employing such models is to minimize the amount of prediction error, one step or several steps ahead.…”

“…where it is assumed that α = 0.2, β = 0.1, and τ = 17 (Althelaya et al 2021). Note that if τ ≥ 17, then there is a chaotic time series (the values of α and β are based on Mackey-Glass equation).…”

Remote Sensing Image Fusion Algorithm Based on Two-Stream Fusion Network and Residual Channel Attention Mechanism

“…Modulation Transfer Function-Generalized Laplacian Pyramid (MTF_GLP) [17] is also proposed. This kind of method with high computational complexity prones to the space distortion [18].…”

Stock Market Prediction Based on Time-frequency Analysis and Convolutional Neural Network

“…Wavelet transform (WT) is combined with stacked autoencoder (SAE) and LSTM to generate a hybrid model WT-SAE-LSTM for U.S. electricity prices forecasting [13]. A stock prices forecasting model combines the empirical wavelet transforms (EWT) with deep RNN is effective when evaluated on the S&P500 stock index and Mackey-Glass time series [14].…”

“…Comparing previous works combining frequency decomposition with neural networks in financial time series forecasting in [11][12][13][14], the contributions of our model combing time-frequency analysis with CNN considering frequency influence are as follows:…”