Sahakian, ValerieKlimasewski, Alexis2021-04-272021-04-272021-04-27https://hdl.handle.net/1794/26187Traditional, empirical ground-motion models (GMMs) are developed by prescribing a functional form between predictive parameters and ground-motion intensity measures. Machine learning techniques may serve as a fully data-driven alternative to regression techniques as they do not require explicitly defining these relationships; however, there are few studies that assess performance of the methods side-by-side. We compare these two approaches: a mixed-effects maximum-likelihood (MEML) model, and a feed-forward artificial neural network (ANN). We develop both models on the same dataset from Southern California and test on the 2019 Ridgecrest sequence to examine model portability. We find that with our small set of input parameters, the ANN shows more site-specific predictions than the MEML model and performs better than their corresponding MEML model when applied “blind” to our testing dataset.en-USAll Rights Reserved.SeismologyElectronic Thesis or Dissertation