WEST Computer Science: this team of scientist/engineer educators designed and taught an inquiry lab activity through ISEE’s Professional Development Program

L-R: Suzanne da Camara, Grace Lin, Ivo Jimenez

Venue: WEST Computer Science 

Team members:

Grace Lin (Design Team Lead): Data Scientist, authors.me

Ivo Jimenez: Graduate Student, Computer Science, UC Santa Cruz

Suzanne da Camara: Graduate Student, Computer Science/Human Computer Interaction, UC Santa Cruz

Audience: 25 undergraduates

Activity Name: Machine Learning with Decision Trees

Learning Goals (Author Grace Lin): 

Our main content goal was to have learners use the concept of a decision tree to create an algorithm that produces a model to solve a predictive problem and maximizes the prediction accuracy. More specifcally this asks learners to 1) understand how to create a decision tree for a specific dataset, 2) how to create a decision tree in general, and 3) how to maximize the prediction accuracy of the model.

One aspect of our design that enhanced learners' content knowledge was the randomly assigned small group formation based around pair programming (https://en.wikipedia.org/wiki/Pair_programming) and peer teaching (http://www.opencolleges.edu.au/informed/features/peer-teaching/). Small groups allow each member to contribute more evenly, and randomly assign people together enforces students to learn to cooperate with strangers.

One aspect of our design that can be improved is to perhaps provide a smaller dataset at the beginning, rather than going straight to a large one. With the smaller dataset it is easier to navigate and get used to the R script that we've provided. It is also possible to use the small dataset throughout the entire inquiry and still learn the concepts, but this needs more thought because small datasets might be too "easy".

Funding for this team provided by: ISEE (UCSC)