Signal Processing Electrical Engineering
Resume:
Abraham Thomas
Current Address:
College Park, MD 20740
Email Address:
*******@***.***.***
Objective:
To contribute and grow as a member of an organization involved
in innovative Research, Product Design and Development in areas
related to Signal Processing, Communications, Estimation and/or
Stochastic Adaptive Control.
Education:
Ph.D. in Electrical Engineering
Expected Graduation: December 1998
University of Maryland, College Park, MD
GPA: 4.0/4.0
Major: Communications, Minor: Computers
Thesis Title: Learning Algorithms for Markov Decision Processes
M.E. in Systems Science and Automation
Graduation: January 1990
Indian Institute of Science, Bangalore, India
GPA: 7.5/8.0
Project Title: Some New Signal Processing Algorithms for ECG Analysis
B.Tech. in Applied Electronics and Instrumentation
Graduation: April 1988
College of Engineering at Trivandrum, University of Kerala, India
Grade: 87.3 %
Experience:
Graduate Research Assistant, Institute of Systems Research
Univ. of Maryland, College Park, MD, August 1995 - present
Graduate Fellow, Institute of Systems Research
Univ. of Maryland, College Park, MD, August 1992 - August 1995
Performed research into the Adaptive Control of Markov Decision
Processes. Extended results to the Partially Observable Markov
Decision Problem (POMDP). Obtained bounds on the Lipschitz
continuity of the value function for POMDPs and devised a scheme
to obtain computationally feasible approximations to the optimal
decision problem. Proved convergence results for a general
Stochastic Approximation Algorithm under weaker assumptions.
The results of the research have direct application to the
fields of operations research, queueing control, and cellular
network handoffs.
Project Associate, Department of Electrical Engineering
Indian Institute of Science, Bangalore, India, January 1991 - July 1992
Designed and developed signal processing algorithms for the
automated identification and classification of relevant
morphological characteristics in Electro Cardiogram for arrhythmia
analysis as part of a multidisciplinary team developing ECG
monitoring tools.
Member (Research Staff), Central Research Labs
Bharat Electronics Ltd., Bangalore, India, March 1990 - June 1990
Conducted analysis and simulation studies to optimize Pre-Beamformer
Automatic Gain Control to identify Active Sonar Returns.
Relevant Graduate Courses:
Linear System Theory, Estimation and Detection, Stochastic Control,
Information Theory, Discrete Event Dynamical Systems,
Parallel Algorithms, Digital Computer Design, Operating Systems,
Real Analysis, Probability Theory, Stochastic Processes, Robotics,
Pattern Recognition, Artificial Intelligence, Computer Vision,
Digital Signal Processing, Optimal Signal Processing,
Computational Methods of Non-linear Optimization
Computer Skills:
Languages : C, Pascal, FORTRAN, LISP
O.S. Environments : UNIX, DOS, X-Windows
Application Software : MATLAB, Island, LaTex
Honors/Activities:
Graduate Fellow, Institute for Systems Research, University of Maryland,
College Park, MD, August 1992 - August 1995
Graduated first in Bachelor of Technology, University of Kerala, India
April 1988 (First Class with Distinction)
Graduated second in Master of Engineering (Systems Science and Automation),
Indian Institute of Science, Bangalore, India
January 1990 (First Class with Distinction)
Reviewer for the IEEE Transactions of Automatic Control.
Professional Memberships:
IEEE, ACM, SIAM, INFORMS
Publications:
A. Thomas and I. S. N. Murthy, R-R Interval Determination in ECGs,
BioMedical Instrumentation Conference, Institution of Engineers (India),
Feb. 1990
A. Thomas and J. V. Avadhanalu, Simulation of Active Sonar Returns and
Pre-Beamformer Dynamic Range Compression, Technical Report, Central
Research Labs, BEL, Bangalore, India, June 1990
A. Thomas and I. S. N. Murthy, Detection of P, QRS and T Wave
components in ECG, Technical Report, Bio-Medical Lab, Dept. of
Electrical Engineering, IISc, Bangalore, India, Aug. 1991
A. Thomas, I. S. N. Murthy and K. Rajagopal, A Pattern Matching
Approach to Classification of QRS Morphologies, Technical Report,
Bio-Medical Lab, Dept. of Electrical Engineering, IISc, Bangalore,
India, June 1992
A. Thomas and S. I. Marcus, Computationally Efficient Schemes for
Partially Observable Markov Decision Processes with Error Bounds,
Proc. 1996 Conference on Information Sciences and Systems, Princeton
Univ., Princeton, NJ, March 1996
A. Thomas and S. I. Marcus, Reinforcement Learning in MDP Framework
using Temporal Difference Schemes, Proc. 1997 Conference on Information
Sciences and Systems, Johns Hopkins Univ., Baltimore, MD, March 1997
A. Thomas and S. I. Marcus, Reinforcement Learning for MDPs using
Temporal Difference Schemes, Proc. 36th IEEE Conference on Decision and
Control, San Diego, CA, December 1997
A. Thomas and S. I. Marcus, Analysis of a Stochastic Approximation
Algorithm, (in preparation)
A. Thomas and S. I. Marcus, Temporal Difference Schemes for Partially
Observable MDPs, (in preparation)
References:
Available upon request
Contact this candidate