ABSTRACT
As Machine Learning (ML) is becoming omnipresent, it is
important to popularize ML knowledge starting in high
school. At this educational stage ML is typically taught by
teaching students to develop ML models for image
classification adopting visual environments. For this
purpose a user-friendly visual layer of Jupyter Notebook
was developed, that provides comprehensive support for all
process steps involved in developing image classification
models. Results of a preliminary evaluation with teachers
and students provide a first indication that it can be a useful,
functional, performance-efficient, and usable support for
teaching the development of image classification models to
novices and ease the transition to conventional
environments later on