In this study, we introduce an optimal rolling force prediction model for H-shaped steel during the hot-rolling process, employing multiple machine learning methods. The traditional rolling force prediction model often relies on simplistic empirical formulas that fail to account for the complex and variable shape of H-shaped steel, leading to potential product defects. To address this limitation and enhance the effectiveness of traditional prediction models, we propose a novel integrated machine learning approach, the Particle Swarm Optimization-Least Squares Support Vector Machine-AdaBoost (PSO-LSSVM-A) model, for predicting the optimal rolling force applied during the hot rolling of H-shaped steel. Initially, the geometric and physical parameters of H-shaped steel during the hot-rolling process are acquired through sensor-based experimental measurements. To ensure data integrity, the Isolation Forest (IForest)algorithm is employed to identify and eliminate outliers. Subsequently, the PSO algorithm is utilized to optimize the calculation parameters involved in the LSSVM modeling process. Finally, the adaptive AdaBoost algorithm, combined with a weight allocation scheme, is integrated to further enhance the prediction effectiveness of the overall model. Through comparisons, we found that the proposed PSO-LSSVM-A model exhibits superior accuracy and stability in predicting the rolling force of H-shaped steel webs during the hot-rolling manufacturing process. The proposed PSO-LSSVM-A model not only considers the intricate geometry and physical characteristics of H-shaped steel but also leverages the optimization capabilities of the PSO algorithm and the ensemble learning power of the AdaBoost algorithm to deliver robust and reliable predictions. By bridging the gap between traditional empirical and advanced machine learning methods, we developed the model representing a significant advancement in optimizing the hot-rolling process for H-shaped steel, ultimately leading to enhanced productivity and product quality, and extended product life in the manufacturing industry.