Battle damage assessment based on an improved Kullback-Leibler divergence sparse autoencoder
The nodes number of the hidden layer in a deep learning network is quite difficult to determine with traditional methods. To solve this problem, an improved Kullback-Leibler divergence sparse autoencoder (KL-SAE) is proposed in this paper, which can be applied to battle damage assessment (BDA). This method can select automatically the hidden layer feature which contributes most to data reconstruction, and abandon the hidden layer feature which contributes least. Therefore, the structure of the network can be modified. In addition, the method can select automatically hidden layer feature without loss of the network prediction accuracy and increase the computation speed. Experiments on University of California-Irvine (UCI) data sets and BDA for battle damage data demonstrate that the method outperforms other reference data-driven methods. The following results can be found from this paper. First, the improved KL-SAE regression network can guarantee the prediction accuracy and increase the speed of training networks and prediction. Second, the proposed network can select automatically hidden layer effective feature and modify the structure of the network by optimizing the nodes number of the hidden layer.
Key wordsBattle damage assessment Improved Kullback-Leibler divergence sparse autoencoder Structural optimization Feature selection
CLC numberTP391.4 E917
Unable to display preview. Download preview PDF.
- Cao, S.C., Zhang, F., 2014. Review of battle damage assessment. Mil. Econ. Res., (8):53–56 (in Chinese).Google Scholar
- Chen, X., Li, L., Liu, D., 2011. Battle damage level prediction on fuzzy theory and Bayesian method. IEEE Conf. on Robotics, Automation and Mechatronics, p.295–298. https://doi.org/10.1109/RAMECH.2011.6070499Google Scholar
- Jiang, N., Rong, W.G., Peng, B.L., et al., 2015. An empirical analysis of different sparse penalties for autoencoder in unsupervised feature learning. Int. Joint Conf. on Neural Networks, p.1–8. https://doi.org/10.1109/IJCNN.2015.7280568Google Scholar
- Li, C.H., Huang, J., 2014. The application of Bayesian network in battle damage assessment. IEEE Int. Conf. on Software Engineering and Service Science, p.529–532. https://doi.org/10.1109/ICSESS.2014.6933622Google Scholar
- Ma, X.M., Ding, P., Yan, W.D., 2016. Warship-damage assessment based on Bayesian networks. Ordnance Ind. Autom., 35(6):72–75 (in Chinese). https://doi.org/10.7690/bgzdh.2016.06.017Google Scholar
- Ma, Z.J., Shi, Q., Li, B., 2007. Battle damage assessment based on Bayesian network. 8th ACIS Int. Conf. on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Computing, p.388–391. https://doi.org/10.1109/SNPD.2007.421Google Scholar
- Rifai, S., Vincent, P., Muller, X., et al., 2011. Contractive auto-encoders: explicit invariance during feature extraction. 28th Int. Conf. on Machine Learning, p.833–840.Google Scholar
- Wen, M.F., Hu, C., Liu, W.R., 2016. Heterogeneous multimodal object recognition method based on deep learning. J. Cent. South Univ. (Sci. Technol.), 47(5):1580–1586 (in Chinese). https://doi.org/10.11817/j.issn.1672-7207.2016.05.018Google Scholar
- Yong, L.Y., 2004. Modeling in Battle Damage Based on Multi-agent. MS Thesis, Harbin University of Science and Technology, Harbin, China (in Chinese).Google Scholar
- Zhang, C., Shi, Q., Liu, T.L., et al., 2012. Study on battle damage level prediction using hybrid-learning algorithm. 4th Int. Conf. on Computational and Information Sciences, p.65–68. https://doi.org/10.1109/ICCIS.2012.298Google Scholar
- Zhao, Z.Y., Li, Y.X., Yu, F., et al., 2015. Improved deep learning algorithm based on extreme learning machine. Comput. Eng. Des., 36(4):1022–1026 (in Chinese). https://doi.org/10.16208/j.issn1000-7024.2015.04.036Google Scholar