Special Session on Geometric Data Analysis
Chair: Fionn Murtagh
Arising in part out of earlier dimensionality reduction work, including in such fields as eugenics, Geometric Data Analysis has now become the central methodology for uncovering the semantics of information, and most often for the visualization and display of data and information. Indeed the visualization and verbalization of data are the keys to how big data analytics can be successfully and reliably addressed. The geometry of information, as well as the topology of information, are closely related, and are integral to any comprehensive approach to contemporary data analytics. The geometry and topology of information, with the methods and techniques for practical, scalable analytics are especially key to data mining and knowledge discovery in data. Thus, in this session, there is particular relevance for unsupervised machine learning, linked to supervised approaches.
This session especially welcomes new results in text and other unstructured data analytics, new perspectives on data analytics arising from the mathematics of geometry and topology, massive data stores or flows, and very high dimensional information spaces. Theoretical innovation and practical case studies, including software systems, are of interest. Fields of application include, but are not limited to, smart cities, cyber-sensor systems, digital humanities and digital society, computational science, health and well-being, and e-discovery for security and forensics.