Uncertainty and forecastability of regional output growth in the UK: Evidence from machine learning

Balcılar, Mehmet; Gabauer, David; Gupta, Rangan; Pierdzioch, Christian

doi:10.1002/for.2851

Cited by 4 publications

(3 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…

\begin{matrix} W = [w_{ij}], i = 1, 2, \dots, p; j = 1, 2, \dots, m \\ Θ = [b_{j}], j = 1, 2, \dots, m \\ {italicneth}_{j} = \sum_{i = 1}^{p} w_{italicij} x_{t - i} + b_{i}, j = 1, 2, \dots, m \end{matrix}

The outputs of the hidden layer nodes

((), {oh}_{j}, j = 1, 2, \dots, m)

are calculated by using logistic activation functions by Equation (3):

{oh}_{j} goodbreak= \frac{1}{1 + \exp ((), goodbreak- {neth}_{j})}, j goodbreak= 1, 2, \dots, m

The output of the pi‐sigma part is calculated with (Alanis et al, 2021).

{output}_{t}^{italicpi - italicsigma} goodbreak= \prod_{j = 1}^{m} {oh}_{j}

The output of the exponential smoothing can be calculated with (Balcilar et al, 2022).

{\overset{truê}{x}}_{t - 1}

is taken as zero in Equation (5) for the calculation for first learning sample.…”

Section: Proposed Method: a New Recurrent Pi‐sigma Artificial Neural ...mentioning

confidence: 99%

“…The output of the exponential smoothing can be calculated with (Balcilar et al, 2022). b x tÀ1 is taken as zero in Equation ( 5) for the calculation for first learning sample.…”

Section: Proposed Method: a New Recurrent Pi-sigma Artificial Neural ...mentioning

confidence: 99%

“…Song et al (2022), Choi andShin (2022), andJiao et al (2022) studies sought solutions to forecasting problems using deep learning methods. Balcilar et al (2022), Yilmaz andYolcu (2022), andLiu et al (2022), on the other hand, conducted research on solving the forecasting problem with machine learning methods.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A new recurrent pi‐sigma artificial neural network inspired by exponential smoothing feedback mechanism

Baş

Eğrioğlu

2022

Journal of Forecasting

View full text Add to dashboard Cite

Pi-sigma artificial neural networks have very good performance for forecasting problems because of their highly nonlinear model structure. Some time series can be forecasted better with the combination of simple and highly nonlinear structures. In this study, the architecture of the pi-sigma artificial neural network is modified by an inspiring exponential smoothing feedback mechanism. The architecture of the proposed neural network automatically balances simple and complex nonlinear model structures. Moreover, a training algorithm is created by using the sine cosine algorithm. The training algorithm has some solutions for overfitting and local optimum problems. The main contribution of the study is to propose a new hybrid recurrent neural network and its training algorithm. In the analysis of the new network's performance, the randomly selected eight subseries of the Financial Times Stock Exchange 100 Index daily time series between 2014 and 2018 are used. The performance of the proposed method is compared with popular deep learning artificial neural networks, pisigma artificial neural networks, and classical exponential smoothing methods. The statistical results show that the proposed method can produce better forecasting results than the others.

show abstract

“…

\begin{matrix} W = [w_{ij}], i = 1, 2, \dots, p; j = 1, 2, \dots, m \\ Θ = [b_{j}], j = 1, 2, \dots, m \\ {italicneth}_{j} = \sum_{i = 1}^{p} w_{italicij} x_{t - i} + b_{i}, j = 1, 2, \dots, m \end{matrix}

The outputs of the hidden layer nodes

((), {oh}_{j}, j = 1, 2, \dots, m)

are calculated by using logistic activation functions by Equation (3):

{oh}_{j} goodbreak= \frac{1}{1 + \exp ((), goodbreak- {neth}_{j})}, j goodbreak= 1, 2, \dots, m

The output of the pi‐sigma part is calculated with (Alanis et al, 2021).

{output}_{t}^{italicpi - italicsigma} goodbreak= \prod_{j = 1}^{m} {oh}_{j}

The output of the exponential smoothing can be calculated with (Balcilar et al, 2022).

{\overset{truê}{x}}_{t - 1}

is taken as zero in Equation (5) for the calculation for first learning sample.…”

Section: Proposed Method: a New Recurrent Pi‐sigma Artificial Neural ...mentioning

confidence: 99%

“…The output of the exponential smoothing can be calculated with (Balcilar et al, 2022). b x tÀ1 is taken as zero in Equation ( 5) for the calculation for first learning sample.…”

Section: Proposed Method: a New Recurrent Pi-sigma Artificial Neural ...mentioning

confidence: 99%

See 1 more Smart Citation

A new recurrent pi‐sigma artificial neural network inspired by exponential smoothing feedback mechanism

Baş

Eğrioğlu

2022

Journal of Forecasting

View full text Add to dashboard Cite

show abstract

Forecasting more than three centuries of economic growth of the United Kingdom: The role of climate risks

Marfatia

Gupta

Aye

et al. 2023

Reference Module in Social Sciences

View full text Add to dashboard Cite

GDP nowcasting: A machine learning and remote sensing data-based approach for Bolivia

Bolivar

2024

Latin American Journal of Central Banking

View full text Add to dashboard Cite

Uncertainty and forecastability of regional output growth in the UK: Evidence from machine learning

Cited by 4 publications

References 60 publications

A new recurrent pi‐sigma artificial neural network inspired by exponential smoothing feedback mechanism

A new recurrent pi‐sigma artificial neural network inspired by exponential smoothing feedback mechanism

Forecasting more than three centuries of economic growth of the United Kingdom: The role of climate risks

GDP nowcasting: A machine learning and remote sensing data-based approach for Bolivia

Contact Info

Product

Resources

About