English title: Enhancing a Social Science Model-building Workflow with Interactive Visualisation
Author(s): Cagatay Turkay - Aidan Slingsby - Kaisa Lahtinen - Sarah Butt - Jason Dykes -
Type: Conference paper/poster
Models can help scientists study and understand phenomena. Such models need to be informed by theory. We report on the early stages of our ongoing study in which we use interactive visualisation to help improve the construction of theory-driven models, by facilitating the exploration of statistical summaries of input variables, compare the quality of alternative models and keep track of the model-building process. Later work will investigate whether machine-learning techniques can be incorporated without compromising the models' theoretical bases.
Conference name: European Symposium on Artificial Neural Networks (ESANN 2016)
Location: Bruges, Belgium
Start date: Apr 27, 2016