An Activity-Based Approach to Teach Machine Learning Using Orange Data Mining for High School Students

Abstract

Machine learning has become an essential component of artificial intelligence. But introducing these concepts to beginners remains challenging due to the programming and mathematical prerequisites of many learning resources. This educational resource presents an activity-based approach for teaching fundamental machine learning concepts to high school students using Orange Data Mining software, a visual programming environment that enables learners to build and evaluate machine learning models without writing code. The instructional package was developed and implemented during the 13th Bahir Dar University Mathcamp to introduce students to artificial intelligence, the machine learning workflow, and the major paradigms of machine learning, including supervised learning, unsupervised learning, reinforcement learning, and deep learning. The hands-on activities focus primarily on supervised and unsupervised learning. Students complete guided exercises on regression using Linear Regression and Neural Networks, classification using Logistic Regression, Neural Networks, and Decision Trees, and clustering using K-Means and Hierarchical Clustering. Each activity includes real-world datasets, model training, prediction on previously unseen data, and interpretation of results. The resource contains presentation slides, datasets, Orange workflow files, student activity sheets, and complete solutions, enabling instructors to adopt or adapt the materials for classroom teaching, workshops, and outreach programs. Classroom implementation demonstrated that students with no prior programming experience were able to understand core machine learning concepts, actively participate in model development, and interpret predictions with confidence. The activity-based design promotes engagement, collaborative learning, and conceptual understanding rather than emphasizing programming syntax. This resource is intended for mathematics, computer science, and STEM educators seeking an accessible introduction to machine learning. By combining visual analytics with structured inquiry-based activities, it provides a practical and reproducible framework for introducing machine learning in secondary schools, first-year university courses, and enrichment programs such as mathematics camps.