GRADIENT RECURRENT NEURAL AND COSINE CROW OPTIMIZED SKILL BASED EMPLOYABILITY PREDICTION IN HIGHER EDUCATION

Abstract

Skill-based employability prediction methods play predominant role for student’s employment.  In this paper, deep learning-based hybrid method of Gradient Recurrent Neural Network and Cosine Crow Search (GRNN-CCS) is introduced for skill-based student employability forecast. Gradient Variance Recurrent Neural Network in the GRNN-CCS method is split into four layers, one input layer, two hidden layers, one output layer to perform intrinsic feature learning for obtaining elevated-level feature representation and therefore predict class of students as employable or less employable in the output layer. Experiments conducted on Student’s Employability Dataset. Results of proposed method can improve accuracy, precision in minimum time and reduces error rate.