The Use of Neural Networks to Predict the Number of Gifted Students

الملخص

In this study, Neural Network shave been used to classify the views into their groups in the presence of some variables that do not follow natural distribution. This is to identify the most important variables that influence admission at the gifted  schools. The criterion of the wrongly rating ratio of viewing has been used as a criterion of the results’ accuracy. The study problem has been the method of distinguishing & classifying gifted students into accepted & not accepted by the National Board of Schools of Gifted Students whose numbers are increasing in their application to be admitted into those kind of schools, which is creating pressure on the National Board of Schools of Gifted Students. Added to this is the multiplicity of statistical methods to process qualitative data & the conditions of using each one of them. the study’s goal is to identify the most important variables influencing the admission to gifted schools in addition to identifying whether the Neural Networks’ method is suitable for processing such data. The study has used the descriptive & analytical inductive methods by analyzing the study data before formulating the results. The study used the program of Statistical Package of Social Science(SPSS version 20) to process the data. One of the most important results of the study was that the said method was moral & provided the importance & effect of the independent variables involved in the analysis. The performance of the Neural Networks was excellent with a good classification of 93%. The most important thing influencing admission was Wexler’s Test. Making use of Neural Networks in classifying data was one of the most important recommendations of the Study. The study suggests the introduction of Genetic Algorithm to the Neural Networks.