On Artificial Neural Networks’ Modelling of Non-Properly Prepared Teachers’ Implication on Students’ Academic Performance, Adopting Noisy Contaminated Optical Character Recognition (OCR)

Abstract

This study tackles a multidisciplinary, fascinating, and important educational phenomenon. The adopted phenomena is directly related to the clarity of the educational environment, which influences the enhancement and illumination of learning/teaching performance. It specifically addresses the severe problem of non-properly prepared teachers having an impact on students' learning performance (achievement) in classrooms. The un favorable amount of improperness is mapped into the well-known communication term signal to noise ratio. This word is abbreviated as SNR or S/N in the context of communication technology, and it evaluates the clarity of the received desired signal across the transmission channel. While bits training to recognize three figures with (T, H, and L) forms using (3X3) retina, the suggested Artificial Neural Network (ANN) model uses a feed forward (FF) structure that follows the Kohonen learning law. After running the suggested realistic simulation program, several interesting results were obtained. Such as the relationship between the value of the learning rate parameter h and the Gaussian additive noise power s learning data submitted by a poorly prepared teacher. Furthermore, the impact of both parameters on students' learning achievement and learning convergence (response time) Herein, this work illustrates specifically the analogy between learning under noisy data environment in Artificial Neural Networks models versus the effect of physical environment on quality of education in classrooms.