Evaluation of Teaching Quality in Higher Mathematics Courses based on Fuzzy-Back Propagation Neural Networks

In this era, teaching quality plays a pivotal role, driving academic excellence, socio-economic growth, and national development, while shaping future generations and fostering global competitiveness. Several methods were proposed to evaluate and enhance teaching quality of higher mathematics which have various limitations like quality issues, lack of security, generalizability, and interpretability. To overcome these problems, a Fuzzy Back-Propagation Neural Network (Fuzzy-BPNN) to evaluate teaching quality of higher mathematics. Initially, a Student Performance (SPD) dataset is considered for evaluation of higher mathematics teaching quality. An Analytical Hierarchy Process (AHP), and Standard normalization are employed for data preprocessing to structure the details of students and teachers, and to analyze the quality of teaching. For feature extraction, a Fast Fourier Series (FFT) is taken to extract relevant features, to enhance scalability and generalizability. The Fuzzy-BPNN is employed for classification to improve overall performance, robustness to noise, and to handle non-linear relationships. The proposed Fuzzy-BPNN model is used to overcome drawbacks of existing system terms of met rices the values are of 98.9% of precision, 99.0% of recall, 99.4% of F1 score and 98.7% of accuracy, when compared to the existing method like Deep Convolutional Neural Network (DCNN-IO).