Publications and Resources
Fast Approximation Algorithms
Ann Arbor Fast Fourier Transform (AAFFT)
This work was partially supported by NSF DMS 0510203.
A full FFT is required by AAFFT. The latest version of FFTW can be found here.
You can download the AAFFT C++ code here. It should compile cleanly with g++.
M. A. Iwen, Anna Gilbert, and Martin Strauss,
Empirical Evaluation of a Sub-Linear Time Sparse DFT Algorithm(.pdf - Adobe Acrobat format),
Communications in Mathematical Sciences, Vol. 5, No. 4, December, 2007.
AAFFT Applications
Richard Baraniuk's lab at Rice University works on AAFFT signal processing applications. My advisor and I collaborated with them on the following paper:
Jason Laska, Sami Kirolos, Yehia Massoud, Richard Baraniuk, Anna Gilbert, Mark Iwen, and Martin Strauss,
Random Sampling for Analog-to-Information Conversion of Wideband Signals (.pdf - Adobe Acrobat format),
IEEE Dallas Circuits and Systems Workshop (DCAS), Dallas, TX, (2006).
- M. A. Iwen,
A Deterministic Sub-linear Time Sparse Fourier Algorithm via Non-adaptive Compressed Sensing Methods(.pdf - Adobe Acrobat format),
ACM-SIAM Symposium on Discrete Algorithms (SODA), San Francisco, CA, (2008).
- M. A. Iwen & C. V. Spencer,
Improved Bounds for a Deterministic Sublinear-Time Sparse Fourier Algorithm(.pdf - Adobe Acrobat format),
Conference on Information Sciences and Systems (CISS), Princeton, NJ, (2008).
Spectral Methods with AAFFT
I've given the following talks using subsets of these slides.
- "Sparse Spectral Methods"
Numerical Analysis of Multiscale Computations Workshop, Banff, January 28-February 2, 2007.
- "Sparse Spectral Methods"
AIM Student Seminar, University of Michigan, January 24, 2007.
- "Empirical Evaluation of a Sub-linear Time Approximate DFT Algorithm"
Sparse Approximation Workshop, Princeton, November 10-12, 2006.
- "Outline of a Sub-linear Time Approximate DFT Algorithm"
AIM Student Seminar, University of Michigan, December 15, 2004.
Bioinformatics
Some Benchmark microarray data sets for AML/ALL, lung, ovarian, and prostate cancer.
More Benchmark microarray data sets for breast cancer, colon cancer,
and others.
M. A. Iwen, W. Lang, and J. Patel,
Scalable Rule-Based Gene Expression Data Classification [Extended Version - BST Classification](.pdf - Adobe Acrobat format),
IEEE International Conference on Data Engineering (ICDE), 2008.
AI Planning
The following papers were written with Professor Amol Mali currently at
the University of Wisconsin Milwaukee.
- Mark Iwen & Amol D. Mali,
DSatz: A Directional SAT Solver for
Planning (.ps -
postscript format or .pdf -
Adobe Acrobat format),
Proceedings of IEEE International Conference on
Tools with Artificial Intelligence (ICTAI), Washington D.C., November 2002,
pp. 199-208.
- Mark Iwen & Amol D. Mali,
Distributed Graphplan (.ps
- postscript format or .pdf
- Adobe Acrobat format),
Proceedings of IEEE International Conference on
Tools with Artificial Intelligence (ICTAI), Washington D.C., November 2002,
pp. 138-145.
- Mark Iwen & Amol D. Mali,
Automatic Problem Decomposition for
Distributed Planning (.ps
- postscript format or .pfd
- Adobe Acrobat format),
Proceedings of the International Conference on
Artificial Intelligence (IC-AI), Vol. 1, Las Vegas, June 2002, pp. 411-417.
- Mark Iwen and Amol D. Mali,
Interaction Graphs for Planning Problem
Decomposition (.ps -
postscript format or .pdf -
Adobe Acrobat format),
Proceedings of international conference on
autonomous agents and multi-agent systems (AAMAS), Italy, 2002, pp. 984-985.
Medical Imaging
I'm currently involved in a project with Dr. Michael Morris to better utilize Raman spectroscopy as a viable medical imaging tool. My advisor and I collaborated with Dr. Morris on the following paper:
M. A. Iwen, G. S. Mandair, M. D. Morris, M. Strauss,
Fast Line-Based Imaging of Small Sample Features (.pdf - Adobe Acrobat format),
IEEE International Conference on Acoustics, Speech, and Signal Processing(ICASSP),
Honolulu, Hawaii, April 2007.
This compendium of NP optimization problems is a useful resource for imaging problems of the kind we discuss above.