Research Articles (Keyur Desai) 
2008

My PhD Dissertation.

2007 (Applied statistics and genomics)

This work will become my PhD dissertation.

KH Desai, JR Deller, JJ McCormick (2007) Tellipsoid: Exploiting inter-gene correlation for improved detection of differential gene expression, submitted to Bioinformatics. [preprint]

KH Desai, JR Deller, JJ McCormick (2007) The distribution of the number of false discoveries for highly correlated null hypotheses. In submission (with revisions) to the Annals of Applied Statstics, preprint, revised version:  main

KH Desai, JR Deller, JJ McCormick (2007) The Distribution of the Number of False Discoveries in DNA Microarray Data. In (SSP'07) IEEE/SP 14th Workshop on Statistical Signal Processing, Madison. Poster

2006

Reading: mostly, math, bio, stats, probabilities, etc.

2004 -- 2005 (Physics, speech recogniftion, biological sequences, etc.)

For a while I worked on a range of research topics until the DNA microarray analysis project came along.

KH Desai1, JR Deller and H Radha(2005) A Learning Framework for Detecting Remote Non-Coding RNA Homologies.  In Proc. 13th Int. Conf. Intelligent Systems for Computational Biol., Detroit.  Poster
(I still have some semi-finished and unpublished stuff from this work.  Because at that time we didn't have a high performance computing center at MSU some of these ideas were tedious to verify since the key statistical model used with noncoding RNA -- stochastic context free grammar first introduced by Chomsky -- has O(N^3) complexity. On a positive note though, this project gave me a decent exposure to biological sequence modeling.)

JR Deller, KH Desai, YP Yang (2005) A decision-theoretic framework for the evaluation of language models used in speech recognizers. In Journal of Natural Language Engineering.
(The idea in this paper, concieved by my mentor Prof. Deller, was quite theoretical and gave me a great exposure to Detection and Estimation.)

K Hoang, KH Desai, SD Mahanti (2005) Charge ordering and self-assembled nanostructures in a fcc Coulomb lattice gas.  In Physical Review B.
(
I did this work while I was working as an RA in the condensed matter group of Prof. Mahanti. I did the statistical and computational work underlying this publication and later on Hoang joined me with his statistical physics expertise to help me get it published. Strucutres discovered in this work became instrumental for a multi-million dollar MURI project on thermoelectric materials and devices. I believe the MURI project is still  in progress at MSU.)


2001 -- 2004 (Signal processing for communications systems)

I started my MS in EE in the fall of 2001. For the first couple of years I was working on Communication circuitry and related signal processing. These are some relics from that old (and really unexciting) world. Surprisingly though, some matrix inversion related stuff that I did in my MS thesis came really handy while  formulating  Tellipsoid.

KH Desai ( 2004)  Advanced Receivers for WCDMA Downlink. MS Thesis Dept. of Electrical and Computer Eng. Michigan State University.
(This MS thesis contains some practical stuff about conjugate gradient and mean square error estimation.)

K Waheed, KH Desai, FM Salem (2003) Blind Multi User Detection in DS-CDMA Systems using Natural Gradient based Symbol Recovery. In Proc. of  4 thInt’l Conference on Independent Component Analysis and Blind Signal Separation, Nara, Japan.

K Waheed, KH Desai, P Seddighrad, FM Salam (2002) A completely integrated, low noise, low power CMOS frequency synthesizer for GSM communications. In Proc. of  45th Midwest Symposium on Circuits and Systems, Tulsa.