Improving computer science education: Teaching neural network modeling to deepen understanding of AI in schools

Abstract

In the context of computer science education, the article highlights a problem area associated with effective teaching of neural network modeling as part of a school computer science course. The emphasis is on analyzing the current state of teaching in various countries, warning about the loss of relevance of a Computer Science course without the inclusion of an artificial intelligence section and limiting access to the best materials on a free basis. The goal of the work is to develop a methodology for teaching 11th graders how to model neural networks using MS Excel and a neurostimulator for a deeper understanding of artificial intelligence. The article describes a three-stage pedagogical experiment, starting with an analysis of the current state of teaching artificial intelligence, developing a methodology and its implementation, and ending with an assessment of effectiveness. Methods used include student surveys, knowledge testing, and neural network modeling project work. It is recommended to use the developed methodology for teaching neural network modeling in schools. Its adaptation to different countries and regions can enrich the educational process. The article is of interest to teachers and specialists in the field of curriculum development in computer science and artificial intelligence.