Machine learning: What is the effect of data size on a sorting activity?

This resource provides a set of videos and a practical investigation aimed at supporting working scientifically in the classroom and relating science to real world experiences. In the first video Professor Brian Cox joins a teacher in a classroom to demonstrate the setting up and carrying out of the experiment. In this paper-based experiment, pupils develop their own rules/algorithms for sorting sweets and aliens and look at the effect that increasing the size of the data set has on the activity. 

In the second video, we visit the Science Museum to discuss online content moderation using AI with Unitary, a company specialising in the use of machine learning systems to reduce the amount of online moderation done by humans. Finally, we visit Cardiff University to meet Professor Pete Burnap, to learn more about cybersecurity and the research being done into technologies used to make our public spaces safer. The second and third videos feature a focus on the skills required to work in these areas and offer young people an insight into which subjects and skills they may wish to pursue later in school and beyond. These could be a useful tool for careers teachers and form tutors.   

This activity is based on an original activity, called All sorted, designed by Oxford Sparks (from the University of Oxford - www.oxfordsparks.ox.ac.uk) who have kindly agreed to its use in the Royal Society Brian Cox School Experiments.

This resource has been provided by the Royal Society.

Classroom experiment

Unitary at the Science Museum

Professor Pete Burnap from Cardiff University