Abstract:

This paper brings teaching experience on simulating the biological process of evolution as a part of explaining the connection between artificial and biological intelligence. Primary use of the simulation is to show complex interactions in, seemingly simple systems. Simulation is based on multiple interacting agents whose action function is determined by individual neural network. Simulation is the use of models to imitate the operation of observed real-world processes or systems, with the purpose of better understanding and/or improving the system. As that, it is often used as a tool in teaching to better explain systems complex in nature. Every year roughly 90 students from Faculty of Science – University of Split, belonging to all STEM fields, enrol winter semester Introduction to AI course. Teachers have to diagnose learning difficulties in order to create best opportunities for learners, with their wishes, plans and ‘inclinations’ [1]. Based on previous teaching experiences and students’ feedback from mentioned university, have demonstrated that students' comprehension of the subject area, such as AI (Artificial Intelligence) – neural networks, was troublesome and that traditional teaching techniques were inadequate to the challenge of the subject matter being taught. The traditional teaching methods include content fragmenting and step-by-step approach [2], but university-level AI courses are first-ever to most of the students and are interconnected e.g., neural networks which draws its roots from biology and the biological definition of neuron. Through this paper, intention is to show how the basic principles of evolution - mutation and natural selection, are introduced to students as a genetic algorithms’ topic within the AI course. This paper aims to show how one concept - evolution - is presented to students, through simulation, to apply biological concepts in artificial intelligence inside one course having heterogeneous group of students.

References:
[1] M. Meyer and B. Hudson, Beyond Fragmentation: Didactics, Learning and Teaching in Europe. Verlag Barbara Budrich, 2011.
[2] C. Graham and P. Longchamps, Transformative Education: A Showcase of Sustainable and Integrative Active Learning. Taylor & Francis, 2022.

Keywords:

AI, teaching, simulation, neurons, neural network, evolution, genetic algorithms.