Effective long short-term memory with fruit fly optimization algorithm for time series forecasting
- PDF / 2,115,566 Bytes
- 21 Pages / 595.276 x 790.866 pts Page_size
- 46 Downloads / 289 Views
(0123456789().,-volV)(0123456789(). ,- volV)
METHODOLOGIES AND APPLICATION
Effective long short-term memory with fruit fly optimization algorithm for time series forecasting Lu Peng1 • Qing Zhu2 • Sheng-Xiang Lv1 • Lin Wang1
Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract A number of recent studies have adopted long short-term memory (LSTM) in extensive applications, such as handwriting recognition and time series prediction, with considerable success. However, the parameters of LSTM have greatly influenced its accuracy and performance. In this study, LSTM with fruit fly optimization algorithm (FOA), called FOALSTM, is designed to solve time series problems. As a novel intelligent algorithm, FOA is applied to decide on the optimal hyper-parameter of LSTM. Experiments under the NN3 time series, three comparative experiments and the monthly energy consumption of the USA are conducted to verify the effectiveness of the FOA-LSTM model. The results indicate that the symmetric mean absolute percentage error (SMAPE) is reduced by up to 11.44% in the last 11 monthly series in the NN3 dataset. Four comparative experiments and the real-life series verify further that the FOA-LSTM model obtains a better result compared with other forecasting models. Keywords Long short-term memory Fruit fly optimization algorithm Time series forecasting
1 Introduction 1.1 Time series prediction based on artificial intelligence One essential research goal of prediction is to forecast the future fluctuation and trend using the history data or the observed phenomenon (Nguyen and Nova´k 2019). The time series data contain sufficient information (Nova´k 2018) and time series prediction has been the subject of
Communicated by V. Loia. & Lin Wang [email protected] Lu Peng [email protected] Qing Zhu [email protected] Sheng-Xiang Lv [email protected] 1
School of Management, Huazhong University of Science and Technology, Wuhan 430074, China
2
International Business School, Shaanxi Normal University, Xi’an 710000, China
considerable studies. For instance, time series prediction has been used for tourism forecasting (Li et al. 2017), sales forecasting (Jimnez et al. 2017), and energy consumption prediction (Zeng et al. 2017; Hu et al. 2020a). These applications in real life have showed the importance of time series prediction, and a large number of methods have been developed for this task. Traditional methods, such as autoregressive integrated moving average (ARIMA) (Hyndman and Khandakar 2008) and exponential smoothing models from innovation state space (ETS) (Hyndman and Khandakar 2008), do not have high-quality information extraction and learning ability. Other widely used time series prediction methods are the artificial neural networks (ANNs), which have been employed in a number of fields, including energy (Zong and Roper 2009; Hu et al. 2020b), economy (Bennett et al. 2013), and environment (Feng et al. 2015). The advantages of ANNs are their massive parallel computing and learning abilities and th
Data Loading...
