In an article published in the journal Nature, researchers introduced a multi-feature fuzzy evaluation model utilizing artificial intelligence (MFEM-AI) to address challenges in assessing physical education teaching methods in colleges and universities (PETCU).
Study: MFEM-AI Revolutionizes Physical Education Teaching in Higher Education. Image credit: Drazen Zigic/Shutterstock
Employing the enhanced cuckoo search optimization algorithm, the framework considered three evaluation perspectives, achieved high scores in various assessment categories (ranging from 87.36% to 97.01%) and demonstrated enhanced efficiency in teaching methods compared to traditional approaches, contributing to advancements in pedagogical practices in the field.
Background
The integration of AI in PETCU has emerged as a promising solution for enhancing the assessment of motor and cognitive functions. The complexity of these processes and the multitude of parameters involved make AI a valuable tool in tracking and identifying changes in both healthy and unwell individuals, including deaf athletes. AI-based performance analysis has become increasingly prevalent in evaluating cognitive and motor skills, offering faster insights into data associations and strategies, particularly when dealing with ambiguous or imperfect data from diverse groups.
Previous research in PETCU evaluation has explored various techniques, including Internet of Things (IoT)-based evaluations, fuzzy systems, machine vision, and AI-driven predictive models. However, these approaches often faced challenges related to data precision, technical complexity, and the need for extensive input. Some studies focused on specific aspects such as basketball skills or art education, while others utilized techniques like neural networks, clustering, and deep learning for assessment.
Despite these efforts, gaps existed in creating a comprehensive, efficient, and precise evaluation framework that considers the diverse dynamics of teaching, student participation, and overall teaching quality in PETCU. This research aimed to fill these gaps by proposing an MFEM-AI. The model integrated natural language, fuzzy control instructions, and an enhanced cuckoo search optimization (ECSO) algorithm to address challenges related to intuitive communication, accommodating imprecise information, and enhancing teaching assessment efficiency in PETCU.
By considering expert input and student dynamics and employing a streamlined system architecture, the MFEM-AI consistently achieved high scores across various assessment categories. This innovative approach provided a comprehensive and effective framework for evaluating and improving PETCU, contributing to advancing pedagogical practices in the field.
Proposed MFEM-AI
The proposed study introduced an MFEM-AI for assessing the quality of PETCU. This model integrated natural language with machine learning, employing fuzzy control instructions to evaluate various aspects of PE, including students' motor abilities like strength, endurance, speed, flexibility, coordination, agility, and balance. The system architecture consisted of three layers: data management (feature layer), indication layer, and AI evaluation layer (target layer). The ECSO algorithm was crucial in fine-tuning fuzzy parameters and optimizing the assessment model for intelligent evaluations.
The evaluation process involved data acquisition and preprocessing, a multi-feature assessment focusing on students' motor abilities, and a three-layered architecture utilizing ECSO for optimization. Traditional assessment methods were compared, highlighting MFEM-AI's superiority in evaluating teaching quality across various categories, such as skill performance, learning progress, physical fitness, participation rate, student satisfaction, and overall teaching efficiency. The ECSO algorithm's workflow was outlined, detailing the stages from data analysis and partitioning to normalization and initialization.
The authors utilized joint nerve-tailored optimization and K-fold cross-validation to optimize AI model parameters effectively. They emphasized the significance of the ECSO algorithm in optimizing parameters related to mobility mechanisms, contributing to a comprehensive evaluation of students' physical abilities in PE. The evaluation system considered three perspectives: management, instructor, and student, with sub-level assessments covering areas like professional competencies, course comprehensiveness, and the impact of PE instruction. The researchers concluded by discussing the ECSO algorithm's features, learning factors, maximum and minimum speeds, and termination criteria. The ECSO algorithm enhanced the cuckoo search (CS) algorithm, providing a more efficient optimization process.
Simulation results and finding
The researchers introduced the MFEM-AI model for PETCU, employing MATLAB fuzzy logic toolbox and the physical activity promotion dataset. The toolbox enabled the creation and analysis of fuzzy logic systems, while the dataset included accelerometer and heart rate data from various physical activities. The authors compared MFEM-AI with traditional models using the ECSO algorithm and evaluated its impact on skill performance, learning progress, physical fitness, participation rate, student satisfaction, and teaching efficiency.
The results demonstrated that MFEM-AI consistently outperformed traditional methods across categories, showcasing its efficacy. Motor skill performance analysis revealed a significant enhancement in MFEM-AI, surpassing other methods by 2.89%, and fitness learning progress analysis showed the model's proficiency with a two-to three-percent improvement. Physical fitness evaluation indicated a 9.71% increase compared to the second-best method. Participation rate analysis demonstrated consistent improvement, ranging from 1.12 to 1.57%.
Student satisfaction assessment displayed high scores for MFEM-AI, emphasizing positive learning experiences. The model's real-time feedback enhanced teaching efficiency, providing immediate adjustments to strategies. Statistical analysis supported the significant difference in skill performance between MFEM-AI and traditional models. Comparative analysis showed an average enhancement of 1.5% across categories. Overall, MFEM-AI effectively assessed and enhanced PETCU, offering a comprehensive framework for higher education institutions.
Conclusion
In conclusion, the MFEM-AI model, integrating fuzzy logic and ECSO optimization, offered a structured and effective approach for assessing PE quality. Achieving high scores across diverse evaluation categories, the model demonstrated proficiency in skill performance, learning progress, physical fitness, participation rate, student satisfaction, and teaching efficiency.
Despite challenges in data reliability and model parameters, the MFEM-AI framework, implemented through MATLAB, proved valuable for precise PETCU assessment. Future research should explore the model's dynamic adaptability and long-term impact, possibly utilizing smartphone applications to enhance interest in PE.
Enhancing Explainability in Generative AI: Key Strategies