ABSTRACT

The observation of the reform and progress in English language instruction reveals a discernible pattern characterised by a focus on proficiency, excellence, and expertise. Since the inception of prioritising knowledge transmission, there has been a gradual shift towards recognising the need of inclusive skills training and development. This has further evolved to place a strong emphasis on fostering excellence in English instruction, with a special focus on nurturing artistic proficiency. Therefore, enhancing the development of humanistic traits in college English instruction is not merely a component of English teaching reform, but rather emerges as the primary concern that English educators must confront. The present study introduces a novel approach, namely the swallow swarm optimisation with a stacked autoencoder (SSOA-SAE) model, to investigate the mental health of college students in the context of English language education. The objective of the SSOA-SAE model is to ascertain the approach to fostering the cultivation of the humanistic spirit in the domains of English instruction and mental health. The SSOA-SAE model is implemented by first executing the SAE model in order to perform the classification process effectively. Furthermore, the Single-Source Optimisation Algorithm (SSOA) is employed to successfully adjust the hyperparameters associated with the Stacked Autoencoder (SAE) model. This study investigates a comprehensive series of simulations conducted on the SSOA-SAE model, and afterwards compares the findings with those of other established models. The predictive accuracy of SSOA-SAE in assessing the mental well-being of students enrolled in English courses with a humanistic approach is evaluated to be 99.88%. The simulation results indicated that the SSOA-SAE model demonstrated superiority above contemporary methodologies.