Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review

Published

International Conference on Data Mining (Las Vegas) - 2010. ed. / Robert Stahlbock; Sven F. Crone; Mahmoud Abou-Nasr; Hamid R. Arabnia; Nikolaos Kourentzes; Philippe Lenca; Wolfram-Manfred Lippe; Gary M. Weiss. CSREA Press, 2010. p. 273-279.

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review

Kourentzes, N & Crone, S 2010, Inference for Neural Network Predictive Models with Impulse Interventions. in R Stahlbock, SF Crone, M Abou-Nasr, HR Arabnia, N Kourentzes, P Lenca, W-M Lippe & GM Weiss (eds), *International Conference on Data Mining (Las Vegas) - 2010.* CSREA Press, pp. 273-279.

Kourentzes, N., & Crone, S. (2010). Inference for Neural Network Predictive Models with Impulse Interventions. In R. Stahlbock, S. F. Crone, M. Abou-Nasr, H. R. Arabnia, N. Kourentzes, P. Lenca, W-M. Lippe, & G. M. Weiss (Eds.), *International Conference on Data Mining (Las Vegas) - 2010 *(pp. 273-279). CSREA Press.

Kourentzes N, Crone S. Inference for Neural Network Predictive Models with Impulse Interventions. In Stahlbock R, Crone SF, Abou-Nasr M, Arabnia HR, Kourentzes N, Lenca P, Lippe W-M, Weiss GM, editors, International Conference on Data Mining (Las Vegas) - 2010. CSREA Press. 2010. p. 273-279

@inproceedings{ce2a36f7f387415d95cb148debae0e89,

title = "Inference for Neural Network Predictive Models with Impulse Interventions",

abstract = "Neural Networks (NN) have demonstrated remarkable time series fitting and prediction abilities, outperforming in several applications other methods and particularly linear models, such as dynamic linear regression. However, due to their nature, NNs are not easy to interpret and are often considered as black box models. The importance of each independent variable is hard to estimate and therefore test whether they have significant explanatory power and hence be included in the model or not. This task is very important for several applications, where the effect of each variable has to be identified, such as marketing modelling and analysis, where the effectiveness of different marketing instruments has to be estimated, commonly modelled as impulse interventions. Statistical inference in these cases is sought, hindering the use of NNs. This paper proposes a framework to allow statistical inference of impulse interventions modelled with NNs. The effects of interventions are estimated and tested for statistical significance. Using a Monte Carlo simulation the power of the proposed test is compared with dynamic linear regression models. The power is found to be higher and the estimation of the simulated effects is more accurate. Based on this framework strategies to code multiple impulses with NNs are discussed.",

author = "Nikolaos Kourentzes and Sven Crone",

year = "2010",

language = "English",

isbn = "1-60132-138-4",

pages = "273--279",

editor = "Robert Stahlbock and Crone, {Sven F.} and Abou-Nasr, {Mahmoud } and Arabnia, {Hamid R.} and Nikolaos Kourentzes and Philippe Lenca and Wolfram-Manfred Lippe and Weiss, {Gary M.}",

booktitle = "International Conference on Data Mining (Las Vegas) - 2010",

publisher = "CSREA Press",

}

TY - GEN

T1 - Inference for Neural Network Predictive Models with Impulse Interventions

AU - Kourentzes, Nikolaos

AU - Crone, Sven

PY - 2010

Y1 - 2010

N2 - Neural Networks (NN) have demonstrated remarkable time series fitting and prediction abilities, outperforming in several applications other methods and particularly linear models, such as dynamic linear regression. However, due to their nature, NNs are not easy to interpret and are often considered as black box models. The importance of each independent variable is hard to estimate and therefore test whether they have significant explanatory power and hence be included in the model or not. This task is very important for several applications, where the effect of each variable has to be identified, such as marketing modelling and analysis, where the effectiveness of different marketing instruments has to be estimated, commonly modelled as impulse interventions. Statistical inference in these cases is sought, hindering the use of NNs. This paper proposes a framework to allow statistical inference of impulse interventions modelled with NNs. The effects of interventions are estimated and tested for statistical significance. Using a Monte Carlo simulation the power of the proposed test is compared with dynamic linear regression models. The power is found to be higher and the estimation of the simulated effects is more accurate. Based on this framework strategies to code multiple impulses with NNs are discussed.

AB - Neural Networks (NN) have demonstrated remarkable time series fitting and prediction abilities, outperforming in several applications other methods and particularly linear models, such as dynamic linear regression. However, due to their nature, NNs are not easy to interpret and are often considered as black box models. The importance of each independent variable is hard to estimate and therefore test whether they have significant explanatory power and hence be included in the model or not. This task is very important for several applications, where the effect of each variable has to be identified, such as marketing modelling and analysis, where the effectiveness of different marketing instruments has to be estimated, commonly modelled as impulse interventions. Statistical inference in these cases is sought, hindering the use of NNs. This paper proposes a framework to allow statistical inference of impulse interventions modelled with NNs. The effects of interventions are estimated and tested for statistical significance. Using a Monte Carlo simulation the power of the proposed test is compared with dynamic linear regression models. The power is found to be higher and the estimation of the simulated effects is more accurate. Based on this framework strategies to code multiple impulses with NNs are discussed.

M3 - Conference contribution/Paper

SN - 1-60132-138-4

SP - 273

EP - 279

BT - International Conference on Data Mining (Las Vegas) - 2010

A2 - Stahlbock, Robert

A2 - Crone, Sven F.

A2 - Abou-Nasr, Mahmoud

A2 - Arabnia, Hamid R.

A2 - Kourentzes, Nikolaos

A2 - Lenca, Philippe

A2 - Lippe, Wolfram-Manfred

A2 - Weiss, Gary M.

PB - CSREA Press

ER -