During the WEKA classes with the teachers, each participant worked with their own dataset, exploring different algorithms and modelling techniques. Some decided to change their datasets mid-session to try more relevant or interesting information for themselves.
During the WEKA classes with the teachers, each participant worked with a dataset of their own, exploring different algorithms and modelling techniques. Some chose pre-defined datasets, while others decided to change their datasets mid-session to test with information more relevant or interesting to them. This added an additional layer of personalisation and challenge to the activity.
As they progressed, the teachers were trying out different algorithms such as classification, regression and clustering, evaluating which offered the best results for their data. An important part of the process was to adjust the parameters of the algorithms, which involved experimenting with different settings to improve the performance of the models. Each iteration required interpreting the results, analysing key metrics such as accuracy, error and other statistics, and then deciding what adjustments to make to optimise performance.
As part of the activity, teachers were also required to complete a worksheet recording the algorithms used, the parameters tested and the results obtained. This allowed them to track their progress in detail, compare approaches and reflect on which strategies work best in their particular case.