In recent years, Education plays an essential role in increasing the life quality and the growing Artificial Intelligence (AI) technology are incorporated in higher education institutions. Predicting academic success has become a key focus because strong academic performance not only boosts university rankings but also opens huge opportunities for students’ job. However, modern educational institutions face numerous challenges, including analyzing student performance, high-quality education and identifying future needs. E-learning is a rapidly expanding mode of education that motivates students to participate in online courses. To address these issues, a model named Triangular Trapezoidal Gaussian Membership-based Neuro Fuzzy Network (TTG-DNFN) is devised for predicting student performance. Initially, the data normalization phase acquires the data gathered from the database, which is carried out by decimal scaling. Next, feature selection is performed using the Wavelet-based Euclidean Distance (WED) method. Finally, the TTG-DNFN is used for student performance prediction, where the membership functions of the Deep Neuro Fuzzy Network (DNFN) are modified by incorporating Triangular, Trapezoidal and Gaussian membership functions. The TTG-DNFN has achieved a high value of 94.83%, 94.77%, 94.80%, and 95.76% for the metrics precision, recall, F-measure, and accuracy.