Home > Archive > 2019 > Volume 9 Number 6 (Dec. 2019) >
IJMLC 2019 Vol.9(6): 734-742 ISSN: 2010-3700
DOI: 10.18178/ijmlc.2019.9.6.866

Performance of Deep Neural Network for Tabular Data — A Case Study of Loss Cost Prediction in Fire Insurance

Dian Maharani, Hendri Murfi, and Yudi Satria

Abstract—The factors that influence fire insurance continue to grow and head to the problem of big data. It is necessary to develop a model to predict the loss cost due to fires by examining the state-of-art models which are adaptable to the big data. One of the models is deep learning, which is an extension of the neural network. This model shows good performances for unstructured data such as image and text. In this paper, we examine the deep learning for loss cost prediction in fire insurance whose training data is tabular or structured data. We use one of the deep learning architectures called deep neural network (DNN), which consists of two or more hidden layers. Our simulation shows that DNN gives quite a similar accuracy to the standard shallow learning of the neural network. It means that deep learning does not improve the performance of the standard shallow learning of neural network for the structured or tabular data of loss cost prediction in fire insurance.

Index Terms—Deep learning, deep neural network, tabular data, structured data, fire insurance, loss cost prediction.

The authors are with Universitas Indonesia, Indonesia (Corresponding author: Dian maharani; e-mail: dian.maharani71@sci.ui.ac.id, hendri@sci.ui.ac.id, ysatria@sci.ui.ac.id).

[PDF]

Cite: Dian Maharani, Hendri Murfi, and Yudi Satria, "Performance of Deep Neural Network for Tabular Data — A Case Study of Loss Cost Prediction in Fire Insurance," International Journal of Machine Learning and Computing vol. 9, no. 6, pp. 734-742, 2019.

Copyright © 2019 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).

 

General Information

  • E-ISSN: 2972-368X
  • Abbreviated Title: Int. J. Mach. Learn.
  • Frequency: Quaterly
  • DOI: 10.18178/IJML
  • Editor-in-Chief: Dr. Lin Huang
  • Executive Editor:  Ms. Cherry L. Chen
  • Abstracing/Indexing: Inspec (IET), Google Scholar, Crossref, ProQuest, Electronic Journals LibraryCNKI.
  • E-mail: ijml@ejournal.net


Article Metrics in Dimensions