DBSCAN Graph Clustering

Making use of the ever-growing amount of data available is a vital opportunity for many industries and research labs, both in Canada and the rest of the world. The initial exploratory data analysis phase is when many of the hypotheses and our intuitive understanding of a data set occurs. Thus, it is essential that the methods used during this phase accurately reflect the data, and avoid making any extreme or exaggerated assumptions. In this research project, we propose to develop a clustering algorithm for graph data that can be used for exploratory analysis, based on the popular DBSCAN clustering algorithm for vector data. The project continues a long and fruitful collaboration on the topic of clustering for graphs, and will result in an open source software implementation of the developed algorithm and a report in a leading scientific journal.

Faculty Supervisor:

Pawel Pralat

Student:

Partner:

SGH Warsaw School of Economics

Discipline:

Mathematics

Sector:

Information and Communications Technology

University:

Toronto Metropolitan University

Program: