Combinatorial algorithms for compressed sensing

Graham Cormode, Shanmugavelayutham Muthukrishnan

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    In sparse approximation theory, the fundamental problem is to reconstruct a signal A ∈ ℝn from linear measurements 〈A, ψ〉 with respect to a dictionary of ψi's. Recently, there is focus on the novel direction of Compressed Sensing [1] where the reconstruction can be done with very few - O(k log n) - linear measurements over a modified dictionary if the signal is compressible, that is, its information is concentrated in k coefficients with the original dictionary. In particular, these results [1], [2], [3] prove that there exists a single O(k log n) × n measurement matrix such that any such signal can be reconstructed from these measurements, with error at most O(1) times the worst case error for the class of such signals. Compressed sensing has generated tremendous excitement both because of the sophisticated underlying Mathematics and because of its potential applications. In this paper, we address outstanding open problems in Compressed Sensing. Our main result is an explicit construction of a non-adaptive measurement matrix and the corresponding reconstruction algorithm so that with a number of measurements polynomial in k, log n, 1/ε, we can reconstruct compressible signals. This is the first known polynomial time explicit construction of any such measurement matrix. In addition, our result improves the error guarantee from O(1) to 1 + ε and improves the reconstruction time from poly(n) to poly(k log n). Our second result is a randomized construction of O(k polylog(n)) measurements that work for each signal with high probability and gives per-instance approximation guarantees rather than over the class of all signals. Previous work on Compressed Sensing does not provide such per-instance approximation guarantees; our result improves the best known number of measurements known from prior work in other areas including Learning Theory [4], [5], Streaming algorithms [6], [7], [8] and Complexity Theory [9] for this case. Our approach is combinatorial. In particular, we use two parallel sets of group tests, one to filter and the other to certify and estimate; the resulting algorithms are quite simple to implement.

    Original languageEnglish (US)
    Title of host publication2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings
    Pages198-201
    Number of pages4
    DOIs
    StatePublished - Dec 1 2007
    Event2006 40th Annual Conference on Information Sciences and Systems, CISS 2006 - Princeton, NJ, United States
    Duration: Mar 22 2006Mar 24 2006

    Publication series

    Name2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings

    Other

    Other2006 40th Annual Conference on Information Sciences and Systems, CISS 2006
    CountryUnited States
    CityPrinceton, NJ
    Period3/22/063/24/06

    Fingerprint

    Compressed sensing
    Glossaries
    Polynomials
    Approximation theory

    ASJC Scopus subject areas

    • Computer Science(all)

    Cite this

    Cormode, G., & Muthukrishnan, S. (2007). Combinatorial algorithms for compressed sensing. In 2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings (pp. 198-201). [4067802] (2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings). https://doi.org/10.1109/CISS.2006.286461

    Combinatorial algorithms for compressed sensing. / Cormode, Graham; Muthukrishnan, Shanmugavelayutham.

    2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings. 2007. p. 198-201 4067802 (2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings).

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Cormode, G & Muthukrishnan, S 2007, Combinatorial algorithms for compressed sensing. in 2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings., 4067802, 2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings, pp. 198-201, 2006 40th Annual Conference on Information Sciences and Systems, CISS 2006, Princeton, NJ, United States, 3/22/06. https://doi.org/10.1109/CISS.2006.286461
    Cormode G, Muthukrishnan S. Combinatorial algorithms for compressed sensing. In 2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings. 2007. p. 198-201. 4067802. (2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings). https://doi.org/10.1109/CISS.2006.286461
    Cormode, Graham ; Muthukrishnan, Shanmugavelayutham. / Combinatorial algorithms for compressed sensing. 2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings. 2007. pp. 198-201 (2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings).
    @inproceedings{e2cabbc44c804884b23a626e5c6b866a,
    title = "Combinatorial algorithms for compressed sensing",
    abstract = "In sparse approximation theory, the fundamental problem is to reconstruct a signal A ∈ ℝn from linear measurements 〈A, ψ〉 with respect to a dictionary of ψi's. Recently, there is focus on the novel direction of Compressed Sensing [1] where the reconstruction can be done with very few - O(k log n) - linear measurements over a modified dictionary if the signal is compressible, that is, its information is concentrated in k coefficients with the original dictionary. In particular, these results [1], [2], [3] prove that there exists a single O(k log n) × n measurement matrix such that any such signal can be reconstructed from these measurements, with error at most O(1) times the worst case error for the class of such signals. Compressed sensing has generated tremendous excitement both because of the sophisticated underlying Mathematics and because of its potential applications. In this paper, we address outstanding open problems in Compressed Sensing. Our main result is an explicit construction of a non-adaptive measurement matrix and the corresponding reconstruction algorithm so that with a number of measurements polynomial in k, log n, 1/ε, we can reconstruct compressible signals. This is the first known polynomial time explicit construction of any such measurement matrix. In addition, our result improves the error guarantee from O(1) to 1 + ε and improves the reconstruction time from poly(n) to poly(k log n). Our second result is a randomized construction of O(k polylog(n)) measurements that work for each signal with high probability and gives per-instance approximation guarantees rather than over the class of all signals. Previous work on Compressed Sensing does not provide such per-instance approximation guarantees; our result improves the best known number of measurements known from prior work in other areas including Learning Theory [4], [5], Streaming algorithms [6], [7], [8] and Complexity Theory [9] for this case. Our approach is combinatorial. In particular, we use two parallel sets of group tests, one to filter and the other to certify and estimate; the resulting algorithms are quite simple to implement.",
    author = "Graham Cormode and Shanmugavelayutham Muthukrishnan",
    year = "2007",
    month = "12",
    day = "1",
    doi = "10.1109/CISS.2006.286461",
    language = "English (US)",
    isbn = "1424403502",
    series = "2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings",
    pages = "198--201",
    booktitle = "2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings",

    }

    TY - GEN

    T1 - Combinatorial algorithms for compressed sensing

    AU - Cormode, Graham

    AU - Muthukrishnan, Shanmugavelayutham

    PY - 2007/12/1

    Y1 - 2007/12/1

    N2 - In sparse approximation theory, the fundamental problem is to reconstruct a signal A ∈ ℝn from linear measurements 〈A, ψ〉 with respect to a dictionary of ψi's. Recently, there is focus on the novel direction of Compressed Sensing [1] where the reconstruction can be done with very few - O(k log n) - linear measurements over a modified dictionary if the signal is compressible, that is, its information is concentrated in k coefficients with the original dictionary. In particular, these results [1], [2], [3] prove that there exists a single O(k log n) × n measurement matrix such that any such signal can be reconstructed from these measurements, with error at most O(1) times the worst case error for the class of such signals. Compressed sensing has generated tremendous excitement both because of the sophisticated underlying Mathematics and because of its potential applications. In this paper, we address outstanding open problems in Compressed Sensing. Our main result is an explicit construction of a non-adaptive measurement matrix and the corresponding reconstruction algorithm so that with a number of measurements polynomial in k, log n, 1/ε, we can reconstruct compressible signals. This is the first known polynomial time explicit construction of any such measurement matrix. In addition, our result improves the error guarantee from O(1) to 1 + ε and improves the reconstruction time from poly(n) to poly(k log n). Our second result is a randomized construction of O(k polylog(n)) measurements that work for each signal with high probability and gives per-instance approximation guarantees rather than over the class of all signals. Previous work on Compressed Sensing does not provide such per-instance approximation guarantees; our result improves the best known number of measurements known from prior work in other areas including Learning Theory [4], [5], Streaming algorithms [6], [7], [8] and Complexity Theory [9] for this case. Our approach is combinatorial. In particular, we use two parallel sets of group tests, one to filter and the other to certify and estimate; the resulting algorithms are quite simple to implement.

    AB - In sparse approximation theory, the fundamental problem is to reconstruct a signal A ∈ ℝn from linear measurements 〈A, ψ〉 with respect to a dictionary of ψi's. Recently, there is focus on the novel direction of Compressed Sensing [1] where the reconstruction can be done with very few - O(k log n) - linear measurements over a modified dictionary if the signal is compressible, that is, its information is concentrated in k coefficients with the original dictionary. In particular, these results [1], [2], [3] prove that there exists a single O(k log n) × n measurement matrix such that any such signal can be reconstructed from these measurements, with error at most O(1) times the worst case error for the class of such signals. Compressed sensing has generated tremendous excitement both because of the sophisticated underlying Mathematics and because of its potential applications. In this paper, we address outstanding open problems in Compressed Sensing. Our main result is an explicit construction of a non-adaptive measurement matrix and the corresponding reconstruction algorithm so that with a number of measurements polynomial in k, log n, 1/ε, we can reconstruct compressible signals. This is the first known polynomial time explicit construction of any such measurement matrix. In addition, our result improves the error guarantee from O(1) to 1 + ε and improves the reconstruction time from poly(n) to poly(k log n). Our second result is a randomized construction of O(k polylog(n)) measurements that work for each signal with high probability and gives per-instance approximation guarantees rather than over the class of all signals. Previous work on Compressed Sensing does not provide such per-instance approximation guarantees; our result improves the best known number of measurements known from prior work in other areas including Learning Theory [4], [5], Streaming algorithms [6], [7], [8] and Complexity Theory [9] for this case. Our approach is combinatorial. In particular, we use two parallel sets of group tests, one to filter and the other to certify and estimate; the resulting algorithms are quite simple to implement.

    UR - http://www.scopus.com/inward/record.url?scp=44049103219&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=44049103219&partnerID=8YFLogxK

    U2 - 10.1109/CISS.2006.286461

    DO - 10.1109/CISS.2006.286461

    M3 - Conference contribution

    AN - SCOPUS:44049103219

    SN - 1424403502

    SN - 9781424403509

    T3 - 2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings

    SP - 198

    EP - 201

    BT - 2006 IEEE Conference on Information Sciences and Systems, CISS 2006 - Proceedings

    ER -