United States | Worldwide
Go
IP TelephonyContact CentersUnified CommunicationsCEBP
Lorraine DenbyLorraine Denby
Research Scientist

"Developing and applying innovative statistical techniques to strategically important business problems and transferring their use to others is what I find the most rewarding."
MY RESEARCH
Throughout of my career I have been mostly interested in statistical methodology that is useful in the analysis of large data sets, particularly in visualizing and conveying the results to the nonstatistical community.  For the last five years my main focus has been in the analysis of network data gathered for the purpose of assessing and monitoring a converged network.
PROJECTS
Network data:
  • ExpertNet (EQM): system designed to assess whether a data network can handle VoIP and if not, easily identify the problem areas for the engineer using the system
  • Chatter (currently CAN in the AAN product line): real-time distributed system for monitoring a converged network with real-time notification of problems
  • Studying the relationship of ping and rtp measurements and developing analysis tools for ORBIT data collected on the real Avaya network
PAVE (Performance Analysis and Visualization of Enterprise Workflow): Developing flow graphs for studying the paths that trouble tickets take from creation to closure, methods for determining the low-lying fruit for process improvement and metrics for assessing the impact of these initiatives.
PUBLICATIONS
Over fifty publications in the statistical literature and a number of publications in the data networking research literature:  For example, Scalable Network Assessment for IP Telephony.  B. Karacali and L. Denby and J. Meloche, Proc. of the 2004 IEEE Intl. Conf. on Communications (ICC-04), 2004 describes our "blame attribution" algorithm that automatically pinpoints trouble spots in the network.  A methodological paper is Random Location and Scale Effects: Model Building Methods for a General Class of Models . W. S. Cleveland, L. Denby, and C. Liu (2000). Computing Science and Statistics, 32, 3-10, Interface Foundations of North America 32 describes a new class of models for taking into account rater bias in the analysis of questionnaire data. I have one patent issued and 9 patent applications in network applications.
CONTACT INFORMATION
Phone+1 908.696.5112
letterld_at_avaya_dot_com
Connect
with Avaya