Bharath K Sriperumbudur
Research

Home | Academics | Research | Publications | Links | CV

Broadly, I am interested in many areas/problems in mathematical and computational statistics, some of which are mentioned below.

• Hilbertian metrics and kernels on probability measures

• Regularization in reproducing kernel Hilbert spaces (RKHS) with applications in machine learning and statistics
  



• Empirical process theory, M-estimation, statistical learning theory


• Optimization
• Convex relaxations for sparsity; majorization-minimization algorithms, bi-level/multi-level programming methods in machine learning.
  
• Statistical applications in signal processing
• speech recognition, blind source separation (ICA), linear inverse problems with sparsity constraints

My past research involved working on problems related to speech recognition, image compression, pattern classification etc. So, I am also interested in some of the following areas in signal processing.

• Multi-rate theory and wavelets

• Statistical signal processing; Blind source separation; Speech processing & recognition (Non-uniform speaker normalization)
• Geometric methods in signal and image processing

Last update: October 7, 2008