Output details
11 - Computer Science and Informatics
Robert Gordon University
Machine learning for improved pathological staging of prostate cancer: A performance comparison on a range of classifiers
This is the first application of Bayesian Networks (BN) to prostate cancer stage prediction and the first to replicate Partin's work on the authoritative BAUS UK dataset. Partin tables are used globally so this is significant to urologists. This output was discussed in a recent article on tools for prostate cancer (Nature Reviews Urology doi:10.1038/nrurol.2013.9) and cited as an important example of probabilistic risk assessment in disease diagnosis and management . The BN method outperformed the gold-standard logistic regression technique and many other machine learning techniques.
A related project won ScotlandIS Young Software Engineer of the Year 2012 (http://www.scotlandis.com/news/stories/software-students-collect-a-clutch-of-awards)