The identification of biological markers is essential because of their meaningful role in clinical decision-making. With data mining that leads to biomarker discovery, we may also gain novel insights into the disease mechanism and the molecular processes. We observed a lack of intuitive toolkits with exploratory user interfaces to help domain experts in biomarker discovery and analysis of the results. Intuitive tools with customizable user interfaces and interactive visualizations can significantly accelerate the development and evaluation of potential marker genes. The thesis focuses on survival analysis and examines and constructs methods and software implementations in the Orange data mining toolkit to identify potential survival marker genes.
|
