@inproceedings{discovery10134081, month = {September}, booktitle = {2019 TC304 Student Contest}, address = {Hannover, Germany}, year = {2019}, title = {Application of an Artificial Neural Network for the CPT-based Soil Stratigraphy Classification}, note = {This version is the author accepted manuscript. For information on re-use, please refer to the publisher's terms and conditions.}, publisher = {International Society for Soil Mechanics and Geotechnical Engineering (ISSMGE)}, url = {http://140.112.12.21/issmge/tc304.htm?=10}, author = {Bertelli, S}, abstract = {Subsurface soil profiling is an essential step in a site investigation. The traditional methods for in situ investigations, such as SPT borings and sampling, have been progressively replaced by CPT soundings since they are fast, repeatable, economical and provide continuous parameters of the mechanical behaviour of the soils. However, the derived CPT-based stratigraphy profiles might present noisy thin layers, and its soil type description might not reflect a textural-based classification (i.e. Universal Soil Classification System, USCS). Thus, this paper presents a straightforward artificial neural network (ANN) algorithm, to classify CPT soundings according to the USCS. Data for training the model have been retrieved from SPT-CPT pairs collected after the 2011 Christchurch earthquake in New Zealand. The application of the ANN to case studies show how the method is a cost-effective and time-efficient approach, but more input parameters and data are needed for increasing its performance.}, keywords = {Soil Classification, Deep Learning, Artificial Neural networks, Cone Penetration Test, Data Analysis.} }