My research consists of developing methods involving optimization for classication (clustering) using a graphical framework. The goal is to partition the data sets efficiently and accurately into a predetermined number of clusters, with or without some prior knowledge of the classication of a small percentage of the nodes. The classication problem is essential to machine learning; it occurs in numerous applications, such as medical or genomics data, spam detection, face recognition, financial predictions, and is finding growing use in text mining studies.
I have also formulated some algorithms for image processing. Processing images is also of critical importance since images and other sensing modes are crucial in procedures of a variety of fields such as medicine (i.e. MRI imaging) and engineering.