A Combined Model Based on Recurrent Neural Networks and Graph Convolutional Networks for Financial Time Series Forecasting

Abstract

Accurate and real-time forecasting of the price of oil plays an important role in the world economy. Research interest in forecasting this type of time series has increased considerably in recent decades, since, due to the characteristics of the time series, it was a complicated task with inaccurate results. Concretely, deep learning models such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) have appeared in this field with promising results compared to traditional approaches. To improve the performance of existing networks in time series forecasting, in this work two types of neural networks are brought together, combining the characteristics of a Graph Convolutional Network (GCN) and a Bidirectional Long Short-Term Memory (BiLSTM) network. This is a novel evolution that improves existing results in the literature and provides new possibilities in the analysis of time series. The results confirm a better performance of the combined BiLSTM-GCN approach compared to the BiLSTM and GCN models separately, as well as to the traditional models, with a lower error in all the error metrics used: the Root Mean Squared Error (RMSE), the Mean Squared Error (MSE), the Mean Absolute Percentage Error (MAPE) and the R-squared (R2). These results represent a smaller difference between the result returned by the model and the real value and, therefore, a greater precision in the predictions of this model.

Experimental evaluation of a machine learning approach to improve the reproducibility of network simulations

“…Simulation ticks create data similar to a time series, which requires a more complex model to be able to address the spatial and temporal dependencies. 63–65…”

“…Simulation ticks create data similar to a time series, which requires a more complex model to be able to address the spatial and temporal dependencies. [63][64][65] Generative adversarial networks (GANs) bear considerable resemblance to our current methodology in the sense that the generator creates ''forgeries'' with the aim of making realistic images (simulation model) and the discriminator receives both forgeries and authentic images and aims to tell them apart (binary classifier). 66 This process is empirically driven as the generator can capture the distribution of real samples.…”

Group Method of Data Handling Using Christiano–Fitzgerald Random Walk Filter for Insulator Fault Prediction

“…In the paper of Salem et al [22], the support vector machine was employed to forecast the deterioration of the room-temperature vulcanized coatings on contaminated glass insulators. Time series forecasting has been applied in several fields, for issues related to financial [23], security [24], energy price [25], traffic flow [26], and epidemiology [27], among others. Considering that leakage current is a strong indication that flashovers may occur, evaluating its evolution concerning time series analysis is a promising alternative and will be the focus of this paper.…”

Prediction of US 30-years-treasury-bonds mouvement and trading entry point using Robust 1DCNN-BiLSTM-XGBoost algorithm

“…Recurrent Neural Networks still also used to time predict financial time series movement. In [11], A Lazcano et al proposed a combined model based on Recurrent Neural Networks and Graph Convolutional Networks for crude oil prices prediction. Their proposed approach uses two pre-trained models on crude oil prices and a third model which receives the input and make the predictions.…”