One of the key elements within seismic hazard analysis is the establishment of appropriate ground‐motion models (GMMs), which are used to predict the levels of ground‐motion intensities by considering various parameters (e.g., source, path, and site). Many empirical GMMs were derived on the basis of a predefined linear or nonlinear equation that is heavily dependent on the a priori knowledge of a functional form that varies between the modelers’ choices. To overcome this issue, this study develops a deep neural network (DNN) trained by the recordings from the Pacific Earthquake Engineering Research Center (PEER) Next Generation Attenuation‐West2 Project (NGA‐West2) database. To this end, we collected 20,900 ground motion recordings from the database and randomly split them into the training, validation, and testing datasets. The refined second‐order neuron is proposed to solve the problem, and the Adam optimizer is used to optimize the performance of the model. The prediction errors are evaluated by three performance indicators (i.e., $R2$⁠, root mean square error, mean absolute error), and the predictive results are compared with previous GMMs developed based on the PEER NGA‐West2 database. The between‐event and within‐event standard deviations (SDs) as well as total SDs are calculated and compared. Based on the comparisons, our model maintains consistent performance (e.g., the dependence of predicted intensity measures on seismological and site‐specific parameters) with the compared GMM. Its relatively small total SDs, especially for longer periods, confirm that the proposed model is associated with better predictive power.

Virtual sensing enables estimation of stress in unmeasured locations of a system using a system model, physical sensors and a process model. The system model holds the relationship between the physical sensors and the desired stress response. A process model processes both the physical sensors and the system model to synthesise virtual sensors that ‘measure’ the desired stress response. Thus, virtual sensing enables mapping between the physical sensors (input) and the desired stress response (output). The system model is a mathematical model of the system based on knowledge or data of the system. Here, the data‐driven system model is constructed directly on data analyses for the specific system. In this paper, supervised learning and data‐driven system models are applied to strain estimation of an offshore wind turbine in the dynamic range through a novel use of principal component analysis (PCA); 40 min of training data is used to establish the data‐driven system model that can estimate the dynamic strain response with high precision for 2 months, while the estimated fatigue damage averaged out to −1.76% of the measured strain response.