KNN (K Nearest Neighbors)

Machine Learning

Abstract

This activity focuses on predicting the property of unknown points based on the similarity of K nearest neighbors on a two-dimensional plane. Specifically, in model 1, students will predict the color of an unknown point using KNN classification. In model 2, students will predict the y value of a point using KNN regression. In both two models, students will discuss the choice of the optimal K.

This activity was developed with NSF support through IUSE-1626765. You may request access to this activity via the following link: IntroCS-POGIL Activity Writing Program. 

• Level: Undergraduate
• Setting: Classroom
• Activity Type: Learning Cycle
• Discipline: Computer Science
• Course: Machine Learning
• Keywords: nearest neighbors, classification, regression