Visual Analytics Referred Lab Class Specification
Summer June 2025
1. Overview
This worksheet is the referred assessment for the lab class quizzes earlier in the year. I expect it to take about 10 hours to complete this worksheet.
The output from this work should be
1. a Jupyter notebook containing your code and descriptive text.
The submission deadline is 13:00 on Monday 21st July: create a single zip file for all the files in your submission and submit on Blackboard.
2. Task Details
Download the UCI breast cancer diagnosis dataset from this link https://archive.ics.uci.edu/ml/datasets/Breast+Cancer+Wisconsin+%28Diagnostic%29
Apply PCA, t-SNE, UMAP, and GTM to this dataset and for each method generate a two-dimensional graph with the diagnosis as a classification label used to colour the projected data points. You may use code from the lab classes 7-10 to help write your own solution to this task. You should put your code and supporting text in a Jupyter notebook. For each visualization, you should write one paragraph explaining what the graph tells the viewer about the data and one paragraph that explains how you chose the projection algorithm parameters (e.g. by systematically varying them).
25 marks will be allocated to each technique and the work will be assessed on the systematic application of the algorithm (15 marks each) and the analysis and discussion of the results (10 marks each).