Mathematics Colloquia and Seminars

Return to Colloquia & Seminar listing

Sparse recovery problems arise in many applications ranging from medical imaging to error correction. Consider a d dimensional vector v with at most n (n << d) non-zero components. We collect N << d nonadaptive linear measurements from v and wish to reconstruct v from these measurements. I will first discuss the idea of sparse reconstruction, and then describe our new greedy algorithm, ROMP, which performs this reconstruction uniformly over all such vectors.