Dr. Laura Brown
Associate Professor, Computer Science
research is centered broadly on the application and design of
methods in artificial intelligence and machine learning. This
work spans from the theoretical design of algorithms for
feature selection and learning Bayesian networks, to the
application of methods across domains including clinical
healthcare, biomedicine, power distribution networks, electric
microgrids, and computer systems research.
Dr. Brown's group is investigating parallel machine learning algorithms on wide variety of datasets. This project seeks to better characterize and understand the algorithms. The evaluation looks beyond the standard machine learning metrics of model quality, such as, accuracy, AUC, precision, recall, etc. and incorporates metrics parallel computing community, such as, speed-up.
Recently, Dr. Brown and Dr. Zhenlin Wang have been examining the use of machine learning methods in computer systems research. Specifically, the work looks to better understand, model, and predict performance of applications in heterogeneous (different computer architectures and hardware configurations) data centers. Preliminary work has verified that machine learning methods can be used to create a decision model to select memory virtualization approaches. The current research aims to model co-run interference, when two applications are run on a single chip, and predict this interference for new hardware settings.
For more information, please visit Dr. Brown's website.
Selective Switching Mechanism in Virtual Machines with Support Vector Machines and Transfer Learning
W. Kuang, L. E. Brown, Z. Wang
Machine Learning (2014)
Accurate Preterm Labor Diagnosis Using a CD55-TLR4 Combination Biomarker Model
S. Pratap, L. E. Brown, M. G. Izban, S. Nowicki, B. J. Nowicki
Journal of Biomedical Science and Engineering, vol. 6, p. 253 (2013)
Impact of Interior Permanent Magnet Rotor Design on Signal Injection Based Sensorless Control and Power Conversion
I. P. Brown, G. Y. Sizov, L. E. Brown
IEEE Energy Conversion Congress and Exposition, Denver, CO. September 2013
Detection of Coherent Groups of Generators and the Need for System Separation using Synchrophasor Data
M. Ali, B. A. Mork, L. J. Bohmann, L. E. Brown
IEEE 7th International Power Engineering and Optimization Conference, Langkawi, Malaysia. June 2013
To Feature Space and Back: Identifying Top Weighted Features in Polynomial Support Vector Machines
L. E. Brown, I. Tsamardinos, D. Hardin
Intelligent Data Analysis, vol. 16, p. 551 (2012)