Despite the hype surrounding digital twin technology and its implementation in other fields, the marine industry has published very few successful, full-scale applications of this technology to date. The development of useful digital twin technology within the industry requires fundamental exploration of data fusion techniques in the marine context. Future development of a useful surface platform digital twin that incorporates advanced machine learning techniques requires preliminary experimentation with more basic, well-understood learning models using full-scale, real-world data. The goal of this work was to lay a foundation for addressing the current knowledge gap through fusion of seakeeping predictions with full-scale measured motions using linear least-squares and neural network corrections. Motion data were obtained from an oceanographic research cruise and these correction methods were applied to frequency-domain response predictions. The success of these relatively simplistic correction models provided insight for the next steps of surface platform data fusion algorithm development.