Abstract
This piece of research addressed asystematic approach forevaluation of an interdisciplinary educational challenging phenomenon observed in children's classrooms. Introduced phenomenon is associated with effects of the physical educational environment on learning performance in classrooms. Specifically, it describes the serious problematic issue of overcrowded effect on classrooms' learning / teaching performance. It has been comparatively presented here versus a challenging phenomenal problem of human's selectivity auditory scene analysis. This problem deals with an auditoryperception phenomenon, namely known as Cocktail Party Problem(CPP), Which originated by the process carried out by the auditory system of a human (child) while listening. Commonly, this process experienced as following one speaker(teacher' speech) in the presence of anotherovercrowded classroom noisy effect. Herein, the adopted approach presented comparative study which motivated by insight into auditory perception, which is derived from original Marr's vision theory. Furthermore, introduced proposal for active audition modeling is motivated by analogous active vision processes, such as that observed during Optical Character Recognition(OCR). In nature, observed OCR as well as pattern recognition processes have to be carried out under non ideal environmental learning condition (under effect of noisy data).
Referring to Artificial Neural Network (ANN) modeling context, the two parameters: Learning rate and Gain factor are considered by the presented simulated comparative study. Accordingly, interesting simulation results have been obtained by theendconclusion of this work declaring the interrelationbetween learning rate values versus different noisy levels. As well as, the effect of intrinsic individual children's differences (gain factor values) on selective attention performance is presented.