Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022;9(1):44.
doi: 10.1186/s40537-022-00599-y. Epub 2022 Apr 26.

Time-series analysis with smoothed Convolutional Neural Network

Aji Prasetya Wibawa  1 , Agung Bella Putra Utama  1 , Hakkun Elmunsyah  1 , Utomo Pujianto  1 , Felix Andika Dwiyanto  1 , Leonel Hernandez  2
Affiliations

Time-series analysis with smoothed Convolutional Neural Network

Aji Prasetya Wibawa et al. J Big Data. 2022.

Abstract

CNN originates from image processing and is not commonly known as a forecasting technique in time-series analysis which depends on the quality of input data. One of the methods to improve the quality is by smoothing the data. This study introduces a novel hybrid exponential smoothing using CNN called Smoothed-CNN (S-CNN). The method of combining tactics outperforms the majority of individual solutions in forecasting. The S-CNN was compared with the original CNN method and other forecasting methods such as Multilayer Perceptron (MLP) and Long Short-Term Memory (LSTM). The dataset is a year time-series of daily website visitors. Since there are no special rules for using the number of hidden layers, the Lucas number was used. The results show that S-CNN is better than MLP and LSTM, with the best MSE of 0.012147693 using 76 hidden layers at 80%:20% data composition.

Keywords: CNN; Exponential smoothing; Optimum smoothing factor; Time-series.

PubMed Disclaimer

Conflict of interest statement

Competing interestsNo competing of interests.

Figures

Fig. 1
Fig. 1
Experimental design of Smoothed-CNN (S-CNN) with optimum α
Fig. 2
Fig. 2
Training testing data composition
Fig. 3
Fig. 3
Time-series component of optimum α
Fig. 4
Fig. 4
Sequence of Fibonacci and Lucas numbers
Fig. 5
Fig. 5
The forecasting results of CNN scenario 1
Fig. 6
Fig. 6
The forecasting results of S-CNN scenario 1
Fig. 7
Fig. 7
Comparison of CNN and S-CNN with scenario 1: a MSE; b Processing time
Fig. 8
Fig. 8
The forecasting results of CNN scenario 2
Fig. 9
Fig. 9
The forecasting results of S-CNN scenario 2
Fig. 10
Fig. 10
Comparison of CNN and S-CNN with scenario 2: a MSE; b Processing time
See this image and copyright information in PMC

References

    1. Velázquez JA, Petit T, Lavoie A, Boucher M-A. An evaluation of the Canadian global meteorological ensemble prediction system for short-term hydrological forecasting. Hydrol Earth Syst Sci. 2009;13(11):2221–2231. doi: 10.5194/hess-13-2221-2009. - DOI
    1. Purnawansyah P, Haviluddin H, Alfred R, Gaffar AFO. Network traffic time series performance analysis using statistical methods. Knowl Eng Data Sci. 2017;1(1):1. doi: 10.17977/um018v1i12018p1-7. - DOI
    1. Singh J, Tripathi P. Time series forecasting using back propagation neural networks. Neurocomputing. 2017;7(5):147–159.
    1. Lundervold AS, Lundervold A. An overview of deep learning in medical imaging focusing on MRI. Z Med Phys. 2019;29(2):102–127. doi: 10.1016/j.zemedi.2018.11.002. - DOI - PubMed
    1. Kim BS, Kim TG. Cooperation of simulation and data model for performance analysis of complex systems. Int J Simul Model. 2019;18(4):608–619. doi: 10.2507/IJSIMM18(4)491. - DOI

LinkOut - more resources