Abstract
This paper discusses penalized neural networks to establish a stable neural network model for survey data measured by traditional marketing scales. Interpreting estimated hidden units and weights in a neural network is often challenging because of its non-identifiability. Factor models in social science are a traditional non-identifiable model for analyzing questionnaire measurements. Hence, many studies have proposed identification conditions. Accordingly, we propose penalty functions that represent the equivalent identification conditions in standard factor models to reduce the non-identifiability and instability of neural networks. We apply these penalty functions in the empirical analysis of autoencoders with e-service quality scale data. The proposed method provides an explainable result that is theoretically reasonable in that e-service quality scale. While comparing the penalized autoencoder with traditional factor models, we discuss potential applications and tasks of the proposed method in service marketing research for further exploration.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Note: Figure 2 displays only the relationship between observed variables and factor scores.Additionally, the intercept is assumed to be equal to zero in the usual case.
- 2.
Note: Although the data were gathered from three categories (food and drink, fashion, and daily necessities), we used only one category data (food and drink) due to limitations of space.
References
de Véricourt, F., Perakis, G.: Frontiers in service science: the management of data analytics services and future directions: New Challenges Serv. Sci. 12(4), 121–129 (2020)
Hollebeek, L.D., Sprott, D.E., Brady, M.K.: Rise of the machines? customer engagement in automated service interactions. J. Serv. Res. 24(1), 3–8 (2021)
Huang, M.H., Rust, R.T.: Artificial intelligence in service. J. Serv. Res. 21(2), 155–172 (2018)
Pemer, F.: Enacting professional service work in times of digitalization and potential disruption. J. Serv. Res. 1094670520916801 (2020)
Yoganarasimhan, H.: Search personalization using machine learning. Manag. Sci. 66(3), 1045–1070 (2020)
Fan, F., Xiong, J., Wang, G.: On Interpretability of Artificial Neural Networks: A survey. Preprint at https://arxiv. org/abs/2001.02522 (2020)
Boratto, L., Carta, S., Fenu, G., Saia, R.: Using neural word embeddings to model user behavior and detect user segments. Knowl.-Based Syst. 108, 5–14 (2016)
Giatsoglou, M., Vozalis, M.G., Diamantaras, K., Vakali, A., Sarigiannidis, G., Chatzisavvas, K.C.: Sentiment analysis leveraging emotions and word embeddings. Expert Syst. Appl. 69, 214–224 (2017)
Antun, V., Renna, F., Poon, C., Adcock, B., Hansen, A.C.: On instabilities of deep learning in image reconstruction and the potential costs of AI. Proc. Natl. Acad. Sci. 117(48), 30088–30095 (2020)
Chugh, M., Whigham, P.A., Dick, G.: Stability of word embeddings using word2vec. In: Mitrovic, T., Xue, B., Li, X. (eds.) AI 2018: Advances in Artificial Intelligence. AI 2018. Lecture Notes in Computer Science, vol. 11320. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-03991-2_73
Korkobi, T., Djemel, M., Chtourou, M.: Stability analysis of neural networks-based system identification. Model. Simul. Eng. (2008)
Rudin, C., Carlson, D.: The secrets of machine learning: ten things you wish you had known earlier to be more effective at data analysis. In: Operations Research & Management Science in the Age of Analytics, pp. 44–72, Informs (2019)
Wendlandt, L., Kummerfeld, J.K., Mihalcea, R.: Factors Influencing the Surprising Instability of Word Embeddings, arXiv preprint arXiv:1804.09692 (2018)
Anderson, T.W., Rubin, H.: Statistical inference in factor analysis. In: Proceedings of the Third Berkeley Symposium on Mathematical Statistics and Probability, vol. 5, pp.111–150 (1956)
Bai, J., Ng, S.: Principal components estimation and identification of static factors. J. Econ. 176(1), 18–29 (2013)
Peeters, C.F.: Rotational uniqueness conditions under oblique factor correlation metric. Psychometrika 77, 288–292 (2012)
Loshchilov, I., Hutter, F.: Decoupled Weight Decay Regularization. arXiv preprint arXiv:1711.05101 (2017)
Moody, J., Hanson, S., Krogh, A., Hertz, J.A.: A simple weight decay can improve generalization. Adv. Neural. Inf. Process. Syst. 4, 950–957 (1995)
Blut, M.: E-Service quality: development of a hierarchical model. J. Retail. 92(4), 500–517 (2016)
Goodfellow, Y., Bengio, A.: Courville and Y. Deep Learning, Cambridge MIT press, Bengio (2016)
Hinton, G.E., Salakhutdinov, R.R.: Reducing the dimensionality of data with neural networks. Science 313(5786), 504–507 (2006)
Adachi, K.: Matrix-based Introduction to Multivariate Data Analysis, Singapore, Springer Singapore (2016)
Tipping, M.E., Bishop, C.M.: Probabilistic principal component analysis. J. R. Stat. Soc. Ser. (Stat. Meth.) 61(3), 611–622 (1999)
Sato, T.: Bayesian estimation for identifiable topic models with latent dirichlet allocation, SSRN Electron. J. https://dx.doi.org/https://doi.org/10.2139/ssrn.3721769 (2020). (unpublished)
Baldi, P., Hornik, K.: Neural networks and principal component analysis: learning from examples without local minima. Neural Netw. 2(1), 53–58 (1989)
Bishop, C.M. (ed.): Pattern Recognition and Machine Learning. ISS, Springer, New York (2006). https://doi.org/10.1007/978-0-387-45528-0
Bourlard, H., Kamp, Y.: Auto-association by multilayer perceptrons and singular value decomposition. Biol. Cybern. 59(4), 291–294 (1988)
Sato, T.: Construct validation for a nonlinear measurement model in marketing and consumer behavior research. Data Sci. Serv. Res. Discuss. Pap. 101 (2019) (unpublished)
Sato, T.: Investigating the impacts of customer experience and attribute performances on overall ratings using online review data: nonlinear estimation and visualization with a neural network. Data Sci. Serv. Res. Discuss. Pap. 27(105), 27-62 (2019)
Xu, Z., Frankwick, G.L., Ramirez, E.: Effects of big data analytics and traditional marketing analytics on new product success: a knowledge fusion perspective. J. Bus. Res. 69(5), 1562–1566 (2016)
Acknowledgments
The author gratefully acknowledges the partial support of JSPS KAKENHI Grant Number 20K22125. The author would also like to thank Enago (www.enago.jp) for the English language review. Additionally, the two reviewers involved provided helpful and insightful comments that improved the overall quality of the manuscript, for which the author is thankful.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Sato, T. (2021). Penalizing Neural Network and Autoencoder for the Analysis of Marketing Measurement Scales in Service Marketing Applications. In: Qiu, R., Lyons, K., Chen, W. (eds) AI and Analytics for Smart Cities and Service Systems. ICSS 2021. Lecture Notes in Operations Research. Springer, Cham. https://doi.org/10.1007/978-3-030-90275-9_3
Download citation
DOI: https://doi.org/10.1007/978-3-030-90275-9_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-90274-2
Online ISBN: 978-3-030-90275-9
eBook Packages: Business and ManagementBusiness and Management (R0)