- https://link.springer.com/article/10.1007%2Fs00521-017-3002-z
Final published version

Research output: Contribution to journal › Journal article › peer-review

Published

**Median-Pi artificial neural network for forecasting.** / Egrioglu, Erol; Yolcu, Ufuk; Bas, Eren; Dalar, Ali Zafer.

Research output: Contribution to journal › Journal article › peer-review

Egrioglu, E, Yolcu, U, Bas, E & Dalar, AZ 2019, 'Median-Pi artificial neural network for forecasting', *Neural Computing and Applications*, vol. 31, no. 1, pp. 307-316. https://doi.org/10.1007/s00521-017-3002-z

Egrioglu, E., Yolcu, U., Bas, E., & Dalar, A. Z. (2019). Median-Pi artificial neural network for forecasting. *Neural Computing and Applications*, *31*(1), 307-316. https://doi.org/10.1007/s00521-017-3002-z

Egrioglu E, Yolcu U, Bas E, Dalar AZ. Median-Pi artificial neural network for forecasting. Neural Computing and Applications. 2019 Jan 18;31(1):307-316. https://doi.org/10.1007/s00521-017-3002-z

@article{9cff50460f434561849e729de016986d,

title = "Median-Pi artificial neural network for forecasting",

abstract = "Datasets with outliers can be predicted with robust learning methods or robust artificial neural networks. In robust artificial neural networks, the architectures become robust by using robust statistics as aggregation functions. Median neural network and trimmed mean neural network are two robust artificial neural networks used in the literature. In these robust artificial neural networks, median and trimmed mean statistics are used as aggregation functions. In this study, Median-Pi artificial neural network is proposed as a new robust neural network for the purpose of forecasting. In Median-Pi artificial neural network, median and multiplicative functions are used as aggregation functions. Because of using median, the proposed network can produce good results for data with outliers. The Median-Pi artificial neural network is trained by particle swarm optimization. The performance of the neural network is investigated by using datasets from the International Time Series Forecast Competition 2016 (CIF-2016). The performance of the proposed method in case of outlier is compared to some other artificial neural networks. Median neural network, trimmed mean neural network, Pi-Sigma neural network and the proposed robust network are applied to time series with outlier, and the obtained results are compared. According to application results, the proposed Median-Pi artificial neural network can produce better forecast results than the other network types.",

keywords = "Forecasting, Outliers, Particle swarm optimization, Robust artificial neural networks",

author = "Erol Egrioglu and Ufuk Yolcu and Eren Bas and Dalar, {Ali Zafer}",

year = "2019",

month = jan,

day = "18",

doi = "10.1007/s00521-017-3002-z",

language = "English",

volume = "31",

pages = "307--316",

journal = "Neural Computing and Applications",

issn = "0941-0643",

publisher = "Springer London",

number = "1",

}

TY - JOUR

T1 - Median-Pi artificial neural network for forecasting

AU - Egrioglu, Erol

AU - Yolcu, Ufuk

AU - Bas, Eren

AU - Dalar, Ali Zafer

PY - 2019/1/18

Y1 - 2019/1/18

N2 - Datasets with outliers can be predicted with robust learning methods or robust artificial neural networks. In robust artificial neural networks, the architectures become robust by using robust statistics as aggregation functions. Median neural network and trimmed mean neural network are two robust artificial neural networks used in the literature. In these robust artificial neural networks, median and trimmed mean statistics are used as aggregation functions. In this study, Median-Pi artificial neural network is proposed as a new robust neural network for the purpose of forecasting. In Median-Pi artificial neural network, median and multiplicative functions are used as aggregation functions. Because of using median, the proposed network can produce good results for data with outliers. The Median-Pi artificial neural network is trained by particle swarm optimization. The performance of the neural network is investigated by using datasets from the International Time Series Forecast Competition 2016 (CIF-2016). The performance of the proposed method in case of outlier is compared to some other artificial neural networks. Median neural network, trimmed mean neural network, Pi-Sigma neural network and the proposed robust network are applied to time series with outlier, and the obtained results are compared. According to application results, the proposed Median-Pi artificial neural network can produce better forecast results than the other network types.

AB - Datasets with outliers can be predicted with robust learning methods or robust artificial neural networks. In robust artificial neural networks, the architectures become robust by using robust statistics as aggregation functions. Median neural network and trimmed mean neural network are two robust artificial neural networks used in the literature. In these robust artificial neural networks, median and trimmed mean statistics are used as aggregation functions. In this study, Median-Pi artificial neural network is proposed as a new robust neural network for the purpose of forecasting. In Median-Pi artificial neural network, median and multiplicative functions are used as aggregation functions. Because of using median, the proposed network can produce good results for data with outliers. The Median-Pi artificial neural network is trained by particle swarm optimization. The performance of the neural network is investigated by using datasets from the International Time Series Forecast Competition 2016 (CIF-2016). The performance of the proposed method in case of outlier is compared to some other artificial neural networks. Median neural network, trimmed mean neural network, Pi-Sigma neural network and the proposed robust network are applied to time series with outlier, and the obtained results are compared. According to application results, the proposed Median-Pi artificial neural network can produce better forecast results than the other network types.

KW - Forecasting

KW - Outliers

KW - Particle swarm optimization

KW - Robust artificial neural networks

U2 - 10.1007/s00521-017-3002-z

DO - 10.1007/s00521-017-3002-z

M3 - Journal article

AN - SCOPUS:85019253391

VL - 31

SP - 307

EP - 316

JO - Neural Computing and Applications

JF - Neural Computing and Applications

SN - 0941-0643

IS - 1

ER -