Research Expertise and Interest

data mining, integer programming, discrete optimization, network flow techniques, clustering, image segmentation, machine vision, pattern recognition

Research Description

Dorit Hochbaum is a Chancellor professor in the Department of Industrial Engineering and Operations Research.  Her research interests are in areas of discrete optimization, network flow techniques, data mining, image segmentation, supply chain management and efficient utilization of resources. She did work on approximation algorithms, location problems; on movement of robots; on routing and distribution problems; on feasibility of VLSI designs; on distribution of data bases on computer networks; on clustering problems and on medical imaging, among others. She has contributed to the analysis of heuristics and approximation algorithms in the worst case, and on the average, and to the complexity analysis of algorithms in general, and nonlinear optimization algorithms in particular. Her recent applications work is on problems related to the homeland security with flow based pattern recognition algorithms, analyzing gene expression databases, scheduling and testing, production planning and supply chain streamlining for high tech industries and logistics and planning problems in various industries. Recent theoretical work focuses on particularly efficient techniques using network flow for data mining and image segmentation and for inverse problems, with applications varying from medical prognosis, error correction, medical imaging, nuclear threat detection, financial risk assessment and prediction, to group rankings and decision problems.