Feed conversion ratio (FCR) is a key indicator of pig productivity, but its measurement is labor-intensive and time-consuming. This study aimed to construct a predictive model for cumulative feeding intake (CFI), which could help estimate FCR more efficiently and reduce the time and effort needed for measurements. This study included a total of 987 Yorkshire boars raised in specific pathogen-free environments, with feeding and growth data collected using automatic feeders. The segmented R package and Bayesian ridge regression (BRR) were used to build a predictive model for CFI. The results showed that the optimal body weight range for predicting FCR was 80–110 kg. The BRR model achieved 80% accuracy for CFI prediction, and FCR calculated from predicted CFI showed 81.4% similarity to the corrected FCR. The results clearly demonstrate that even with a limited training dataset, the BRR model has good predictive potential for FCR. The findings of this study could reduce the selection pressure on FCR traits, decrease production costs, and shorten measurement periods, ultimately benefiting the swine industry significantly.