A novel decomposed-ensemble time series forecasting framework: capturing underlying volatility information

TL;DR

This paper introduces a new time series forecasting framework that combines mode decomposition, volatility extraction, and neural networks to improve prediction accuracy for complex data.

Contribution

It integrates Variational Mode Decomposition, GARCH-based volatility modeling, and neural networks into a unified forecasting framework, capturing both data and volatility information.

Findings

Significant reduction in MSE, RMSE, and MAPE compared to baseline models.

Effective capture of underlying volatility improves forecast accuracy.

Demonstrates superior performance on complex time series datasets.

Abstract

Time series forecasting represents a significant and challenging task across various fields. Recently, methods based on mode decomposition have dominated the forecasting of complex time series because of the advantages of capturing local characteristics and extracting intrinsic modes from data. Unfortunately, most models fail to capture the implied volatilities that contain significant information. To enhance the prediction of contemporary diverse and complex time series, we propose a novel time series forecasting paradigm that integrates decomposition with the capability to capture the underlying fluctuation information of the series. In our methodology, we implement the Variational Mode Decomposition algorithm to decompose the time series into K distinct sub-modes. Following this decomposition, we apply the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model to extract the volatility information in these sub-modes. Subsequently, both the numerical data and the volatility information for each sub-mode are harnessed to train a neural network. This network is adept at predicting the information of the sub-modes, and we aggregate the predictions of all sub-modes to generate the final output. By integrating econometric and artificial intelligence methods, and taking into account both the numerical and volatility information of the time series, our proposed framework demonstrates superior performance in time series forecasting, as evidenced by the significant decrease in MSE, RMSE, and MAPE in our comparative experimental results.