You can manage bookmarks using lists, please log in to your user account for this.
Media type:
E-Book
Title:
Forecasting Crude Oil Future Prices from a Sentiment Analysis of Combined News Headlines
:
A Decomposition-Ensemble Based Deep Learning Approach
Description:
As the price of crude oil has nonlinearity, instability, and randomness, capturing its behavior exactly is significantly challenging and results in difficulties in predicting. This paper combines the decomposition-ensemble method, optimized by the seagull algorithm, with a sentiment analysis to handle the problem. First, the cumulative sentiment score sequence is obtained by a sentiment analysis of news headlines data. Second, we extract features from the crude oil future price dataset and decrease the influence of noise. An adaptive signal decomposition approach, namely, ensemble empirical mode decomposition (EEMD), is introduced to decompose the crude oil future price sequence into several intrinsic mode functions and one residual component. Third, a seagull optimization algorithm (SOA) is introduced to tune the hyperparameters of gated recurrent units (GRUs). The optimized GRU model is established to acquire the predicting values of each component integrated with the cumulative sentiment score sequence. Subsequently, multiple linear regression (MLR) is then introduced as the ensemble approach that integrates the forecasting results of each component. The experimental results of daily West Texas Intermediate (WTI) crude oil future and news headlines data from January 4, 2010, to September 17, 2019, validate our proposed decomposition-ensemble approach with different forecasting horizons. This method significantly outperforms others comparison models in terms of the hypothesis test and forecasting accuracy. In addition, the WTI crude oil future prices data set from January 2000 to June 2021 is selected to analyze the effect of black-swan events on crude oil price fluctuations from 21st century