[01] Kernan Mzelikahle, Computer Science Department, National University of Science and Technology, Bulawayo, Zimbabwe. [02] Dumisani John Hlatywayo, Applied Physics Department, National University of Science and Technology, Bulawayo, Zimbabwe. [03] John Trimble, Industrial Engineering Department, Tshwane University of Technology, Tshwane, South Africa.

The BOSOM-LSTM technique is a hybrid neural network capable of conducting generic data analysis. There are a number of factors that generally affect data analysis, and chief among them is the extensibility of an analysis tool. Any tool and/or technique used for analysis is limited in its operation by its own ability to adapt for use in different circumstances. Extensibility of an analysis tool, therefore, implies the ability of a tool to be reconfigured with new configuration parameters in order to suite a new demand. The BOSOM-LSTM technique is a relatively flexible and reconfigurable tool applicable for generic data analysis. In this study, the BOSOM-LSTM was configured for seismic vulnerability analysis using data from Zimbabwe. The objective of the study was to apply the BOSOM-LSTM technique in the assessment of seismic vulnerability of a Zimbabwean city (Mutare City), given a simulated seismic scenario. In this paper, a seismic event was simulated using the VISCO1D software on the data obtained from the East-Southern Africa Rift System. In order to assess seismic vulnerability of public infrastructure and civilian buildings, construction data was obtained from Mutare City Council. This data revealed that construction in the city was based on reinforced concrete material, thus vulnerability of infrastructure in the city could be extrapolated from the compressive strength of reinforced concrete. Results in this paper reveal two significant observations: (1) that the BOSOM-LSTM was successfully configured and used for seismic vulnerability assessment, and (2) that there is significant seismic vulnerability in Mutare City. A conclusion was drawn that the BOSOM-LSTM is applicable in seismic vulnerability assessment. However, a limitation was noted in that the BOSOM-LSTM technique depends on manual parameter tuning techniques, and configuration.

BOSOM-LSTM Technique, Long Short Term Memory, Bat Optimised Self Organising Map, Artificial Neural Networks, Unsupervised Learning, Seismic Vulnerability Assessment

[01] Marfai, M. A. and Njagih, J. K. (2004). Vulnerability Analysis and Risk Assessment for Seismic and Flood Hazard in Turialba City, Costa Rica. International Institute for Geo-Information Sciences and Earth Observation, Enschende Netherlands, pp. 1 – 24. [02] Sandi, H., Pomonis, S., Francis, S., Geogescu, E. S., Mahindra, R. and Borcia, I. S. (2007). Seismic Vulnerability Assessment: Methodological Elements and Applications to the case of Romania. In: Proceedings of the International Symposium on Strong Vrancea Earthquakes and Risk Mitigation, October 4 - 6, 2007. Bucharest, Romania. pp. 328 - 341. [03] Grünthal, G. (1999). Seismic hazard assessment for Central, North and Northwest Europe: GSHAP Region 3. Annali di Geofisica, 42, pp. 999 – 1011. [04] Grünthal, G. and Wahlström, R. (2003). An MW Based Earthquake Catalogue for Central, Northern and Northwestern Europe using a Hierarchy of Magnitude Conversions. Journal of Seismology, 7, pp. 507 – 531. [05] Cornell, C. A. (1968). Engineering Seismic Risk Analysis. Bulletin of the Seismological Society of America, pp. 1583 – 1606. [06] Berse, K. B., Bendimerad, F. and Asami Y. (2011). Beyond geo-spatial technologies: Promoting spatial thinking through local disaster risk management planning. International Journal of Spatial Thinking and Geographic Information Sciences, 21, pp. 73 – 82. [07] Calabrese, A. and Lai, C. G. (2012). Seismic Vulnerability Analysis of Wharf Structures using Artificial Neural Networks. In: 15th World Conference on Earthquake Engineering 2012 (15 WCEE) [08] Gutenberg, B. and Richter, C. (1956). Earthquake Magnitude, Intensity, Energy and Acceleration. Bulletin of the Seismological Society of America, pp. 105-145. [09] Jackson, D. D. (2012). Earthquake Prediction Evaluation Standards applied to the VAN Method. Geophysical Research Letters, 23, pp. 1363 – 1366. [10] Mzelikahle, K., Mapuma, D. J., Hlatywayo, D. J. and Trimble, J., (2017). Optimisation of Self Organising Maps Using the Bat Algorithm. American Journal of Information Science and Computer Engineering, 3 (6), pp. 77-83. [11] Mzelikahle, K., Trimble, J. and Hlatywayo, D. J., (2018). A Hybrid Technique Between BOSOM and LSTM for Data Analysis. International Journal of Mathematics and Computational Science, 4 (4), pp. 128 – 138. [12] Pollitz, F. F. (2007). VISCO1D (version 3). [Online] U. S. Geological Survey. Available From: https://earthquake.usgs. gov/research/software/#VISCO1D [Accessed: 04 November 2018]. [13] ESARSW (2013). Seismological Bulletin Results from Eastern and Southern Africa Regional Workshop. In: ESARS Workshop, 14 – 18 October 2013, Asmara, Eritrea. [14] Joyner, W. B. and Boore, D. M. (1981). Peak Horizontal Acceleration and Velocity from Strong Motion Records Including Records from 1979 Imperial Valley, California earthquake. Bulletin of the Seismological Society of America, pp. 2011 – 2038. [15] Langridge, S., Hartge, E., Clark, R., Arkema, K., Verutes, G., Prahler, E., Stoner-Duncan, S., Revell, D., Caldwell, M., Guerry, A., Ruckelshaus, M., Abeles, A., Coburn, C. and O’Connor, K. (2014). Key Lessons for Incorporating Natural Infrastructure into Regional Climate Adaptation Planning. Journal of Ocean and Coastal Management, 95, pp. 189 – 197. [16] Yeh, I. C. (1998). Modeling of Strength of High-performance Concrete using Artificial Neural Networks. Cement and Concrete Research, 28 (12), pp. 1797 – 1808. [17] Yeh, I. C. (2007). Modeling Slump Flow of Concrete using Second-order Regressions and Artificial Neural Networks. Cement and Concrete Composites, 29 (6), pp. 474 – 480. [18] Ding, X., Li, C., Xu, Y., Li, F. and Zhao, S. (2016). Dataset of Long-term Compressive Strength of Concrete with Manufactured Sand. Data in Brief, Elsevier, 6, pp. 959 – 964. [19] Demuth, H. B., Beale M. H, De Jess, O. and Hagan, M. T. (2014). Neural Network Design. 2nd ed., Martin Hagan, Oklahoma, USA: Oklahoma State University. [20] Graupe, D. (2007). Principles of Artificial Neural Networks. 2nd Ed. New Jersey, USA: World Scientific. [21] Kröse, B. J. and van der Smagt P. (1996). An Introduction to Neural Networks. 8th ed. Department of Computer Systems, University of Amsterdam, Netherlands. [22] Midzi, V., Hlatywayo, D. J., Chapola, L. S., Kebede, F., Atakan, K., Lombe, D. K., Turyomurugyendo, G. and Tugume, F. A., 1999. Seismic hazard assessment in Eastern and Southern Africa. Annals of Geophysics, 42 (6). [23] Shumba, B. T., Hlatywayo, D. J. and Midzi, V. (2009). Focal Mechanism Solution of the 15th March 2008, Nyamandlovu Earthquake. South African Journal of Geology, 112, pp. 381 – 386. [24] Gers, F. A. and Schmidhuber, J. (2001). LSTM Recurrent Networks Learn Simple Context Free and Context Sensitive Languages. IEEE Transactions on Neural Networks, 12 (6), pp. 1333 – 1340. [25] Gers, F. A., Schraudolph, N. N. and Schmidhuber, J. (2002). Learning Precise Timing with LSTM Recurrent Networks. Journal of Machine Learning Research, 3, pp. 115 – 143.