My research interests are in the areas of signal processing and machine learning. I work on a variety of fundamental problems (including boosting, adaptive signal processing, and learning from data) and in a variety of application domains (including fMRI analysis, adaptive control, optimization of queuing systems, and video analysis, annotation, and search). My recent research on function Magnetic Resonance Imaging (fMRI) analysis has centered on developing efficient algorithms for extracting information from large fMRI data sets. For example, in collaboration with neuroscience colleagues, we have developed algorithms for functionally aligning the cortices of multiple subjects using the fMRI data measured during movie viewing. To do so, we used have used two distinct alignment metrics: the correlation of corresponding time series and a metric based on aligning intra-subject functional connectivity. In other recent work we have examined the problem of spatially informed voxel selection from fMRI data, using methods based on spatially regularized boosting and level-set estimation. In ongoing work we are examining the issue of tracking the presence of people in a movie stimulus over time, based on fMRI data collected during the viewing of the stimulus. My research group is also currently investigating problems of signal representation using trees and wavelet constructions, semi-supervised clustering of data, learning patterns in data, manifold regularization and estimation, and online learning. I am an active member of the neuroimaging analysis methods group at Princeton, the recipient of several teaching awards, an IBM faculty development award, and an IEEE best paper award. I am a fellow of the IEEE and a member of the Society for Industrial and Applied Mathematics (SIAM).
Zhen James Xiang and Peter J. Ramadge, “Morphological wavelet transform with adaptive dyadic structures,” IEEE Int. Conf. on Image Processing, Hong Kong, 2010.
Alexander Lorbert and Peter J. Ramadge, “Descent methods for tuning parameter refinement,” Thirteenth AISTATS; 9:469-476, 2010.
Alexander Lorbert, David Eis, Victoria Kostina, David Blei, Peter Ramadge, “Exploiting Covariate Similarity in Sparse Regression via the Pairwise Elastic Net,” Thirteenth AISTATS; 9:477-484, 2010.
Eugene Brevdo and Peter J. Ramadge, “Bridge Detection and Robust Geodesics Estimation via Random Walks.” IEEE ICASSP, Dallas, Texas, March 2010.
Zhen James Xiang and Peter J. Ramadge, “Morphological Wavelets and the Complexity of Dyadic Trees.” IEEE ICASSP, Dallas, Texas, March 2010.
Bryan Conroy, Yongxin Taylor Xi, Peter Ramadge, “A Supervisory Approach to Semi-Supervised Clustering.” IEEE ICASSP, Dallas, Texas, March 2010.
Alexander Lorbert and Peter J. Ramadge, “Level Set Estimation on the Sphere.” IEEE ICASSP, Dallas, Texas, March 2010.
Bryan Conroy, Ben Singer, James Haxby and Peter Ramadge, “fMRI-Based inter-subject cortical alignment using functional connectivity,” NIPS 2009, Dec. 2009, Vancouver, Canada.
Zhen James Xiang, Yongxin Taylor Xi, Uri Hasson and Peter Ramadge, “Boosting with spatial regularization,” NIPS 2009, Dec. 2009, Vancouver, Canada.
Yongxin Taylor Xi and Peter J. Ramadge, ``Using sparse regression to learn effective projections for face recognition,’’ IEEE Int. Conf. Image Processing, Cairo Egypt, Oct., 2009.
M. Sabuncu, B. Singer, Bryan Conroy, P. Ramadge, J. Haxby, “Function-based inter-subject alignment of human cortical anatomy,” Cerebral Cortex, 2009.
Zhen James Xiang and P. J. Ramadge, “Sparse boosting.” ICASSP, Taiwan, April 18-24, 2009.
Yongxin Taylor Xi and P. J. Ramadge, “Separable PCA for image classification.” ICASSP, Taiwan, April 18-24, 2009.
Shannon M. Hughes and Peter J. Ramadge, “Connecting spectral and spring methods for manifold learning.” ICASSP, Taiwan, April 18-24, 2009. (Best Student Paper Award)
Yongxin Taylor Xi, Zhen James Xiang, Peter Ramadge and Robert Schapire “Speed and sparsity of regularized boosting.” AISTATS 2009, April, 2009.