Challenge Based Learning: A Fast Track To Introduce Engineering Students To Data Science

ABSTRACT

This article describes a data science challenge-based learning experience introduced to non-IT second-year engineering students. The methodology proposed in this article was successful in the short introductory course. The students presented well-considered, practical solutions to two challenges of a different nature derived from excellent, quality data processing. The students used free-access databases from Airbnb and Johns Hopkins University to tackle both challenges. Although the students' data analysis methods corresponded more to data analytics, the two student teams incorporated Machine Learning techniques and exceeded our initial expectations. The selection of the programming environment for this experience was a crucial issue addressed in this article. To illustrate the students' work in the course using this methodology, we present a selection of their results, including a new index to measure the degree of herd immunity in a country, relating this index with the possible appearance of a new strain of COVID-19.