Home > Research > Publications & Outputs > A New Bootstrapped Hybrid Artificial Neural Net...

Research

Associated organisational unit

Electronic data

  • ERol Computational Economics_2020_finalRevised paper

    Rights statement: The final publication is available at Springer via http://dx.doi.org/10.1007/s10614-020-10073-7

    Accepted author manuscript, 1.45 MB, PDF document

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License

Links

Text available via DOI:

Keywords

View graph of relations

A New Bootstrapped Hybrid Artificial Neural Network Approach for Time Series Forecasting

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

A New Bootstrapped Hybrid Artificial Neural Network Approach for Time Series Forecasting. / Eğrioğlu, E.; Fildes, R.
In: Computational Economics, Vol. 59, No. 4, 30.04.2022, p. 1355-1383.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Eğrioğlu, E & Fildes, R 2022, 'A New Bootstrapped Hybrid Artificial Neural Network Approach for Time Series Forecasting', Computational Economics, vol. 59, no. 4, pp. 1355-1383. https://doi.org/10.1007/s10614-020-10073-7

APA

Eğrioğlu, E., & Fildes, R. (2022). A New Bootstrapped Hybrid Artificial Neural Network Approach for Time Series Forecasting. Computational Economics, 59(4), 1355-1383. https://doi.org/10.1007/s10614-020-10073-7

Vancouver

Eğrioğlu E, Fildes R. A New Bootstrapped Hybrid Artificial Neural Network Approach for Time Series Forecasting. Computational Economics. 2022 Apr 30;59(4):1355-1383. Epub 2020 Nov 20. doi: 10.1007/s10614-020-10073-7

Author

Eğrioğlu, E. ; Fildes, R. / A New Bootstrapped Hybrid Artificial Neural Network Approach for Time Series Forecasting. In: Computational Economics. 2022 ; Vol. 59, No. 4. pp. 1355-1383.

Bibtex

@article{a8298fd91b6644b39885b81638d9e705,
title = "A New Bootstrapped Hybrid Artificial Neural Network Approach for Time Series Forecasting",
abstract = "In this study, a new bootstrapped hybrid artificial neural network is proposed for forecasting. This new neural network provides input significance, linearity and nonlinearity hypothesis tests in a unique network structure via a residual bootstrap approach. The network has three parts: linear, non-linear and a combination with associated weights and biases. These weights are used to test the input significance, linearity and nonlinearity hypotheses with this new method providing empirical distributions for forecasts and weights. The proposed method employs a bagging approach to obtain forecasts. It is then applied to real-time series including the M4 Competition data set and stock exchange time series where its performance is compared with appropriate benchmark methods including other popular neural networks. The proposed method results are less affected than other neural networks by initial random weights, which means that the results of the proposed method are more stable and precise. The new method provides improvements in forecasting accuracy over the established benchmarks. ",
keywords = "Artificial neural networks, Bootstrap, Deep learning, Forecasting, Input significance, Interval forecast",
author = "E. Eğrioğlu and R. Fildes",
note = "The final publication is available at Springer via http://dx.doi.org/10.1007/s10614-020-10073-7 ",
year = "2022",
month = apr,
day = "30",
doi = "10.1007/s10614-020-10073-7",
language = "English",
volume = "59",
pages = "1355--1383",
journal = "Computational Economics",
issn = "0927-7099",
publisher = "Springer Netherlands",
number = "4",

}

RIS

TY - JOUR

T1 - A New Bootstrapped Hybrid Artificial Neural Network Approach for Time Series Forecasting

AU - Eğrioğlu, E.

AU - Fildes, R.

N1 - The final publication is available at Springer via http://dx.doi.org/10.1007/s10614-020-10073-7

PY - 2022/4/30

Y1 - 2022/4/30

N2 - In this study, a new bootstrapped hybrid artificial neural network is proposed for forecasting. This new neural network provides input significance, linearity and nonlinearity hypothesis tests in a unique network structure via a residual bootstrap approach. The network has three parts: linear, non-linear and a combination with associated weights and biases. These weights are used to test the input significance, linearity and nonlinearity hypotheses with this new method providing empirical distributions for forecasts and weights. The proposed method employs a bagging approach to obtain forecasts. It is then applied to real-time series including the M4 Competition data set and stock exchange time series where its performance is compared with appropriate benchmark methods including other popular neural networks. The proposed method results are less affected than other neural networks by initial random weights, which means that the results of the proposed method are more stable and precise. The new method provides improvements in forecasting accuracy over the established benchmarks.

AB - In this study, a new bootstrapped hybrid artificial neural network is proposed for forecasting. This new neural network provides input significance, linearity and nonlinearity hypothesis tests in a unique network structure via a residual bootstrap approach. The network has three parts: linear, non-linear and a combination with associated weights and biases. These weights are used to test the input significance, linearity and nonlinearity hypotheses with this new method providing empirical distributions for forecasts and weights. The proposed method employs a bagging approach to obtain forecasts. It is then applied to real-time series including the M4 Competition data set and stock exchange time series where its performance is compared with appropriate benchmark methods including other popular neural networks. The proposed method results are less affected than other neural networks by initial random weights, which means that the results of the proposed method are more stable and precise. The new method provides improvements in forecasting accuracy over the established benchmarks.

KW - Artificial neural networks

KW - Bootstrap

KW - Deep learning

KW - Forecasting

KW - Input significance

KW - Interval forecast

U2 - 10.1007/s10614-020-10073-7

DO - 10.1007/s10614-020-10073-7

M3 - Journal article

VL - 59

SP - 1355

EP - 1383

JO - Computational Economics

JF - Computational Economics

SN - 0927-7099

IS - 4

ER -