an:05917979
Zbl 1332.68099
Arora, Sanjeev; Ge, Rong
New algorithms for learning in presence of errors
EN
Aceto, Luca (ed.) et al., Automata, languages and programming. 38th international colloquium, ICALP 2011, Zurich, Switzerland, July 4--8, 2011. Proceedings, Part I. Berlin: Springer (ISBN 978-3-642-22005-0/pbk). Lecture Notes in Computer Science 6755, 403-415 (2011).
2011
a
68Q32 68T05 94A60
Summary: We give new algorithms for a variety of randomly-generated instances of computational problems using a linearization technique that reduces to solving a system of linear equations.
These algorithms are derived in the context of learning with structured noise, a notion introduced in this paper. This notion is best illustrated with the learning parities with noise (LPN) problem -- well-studied in learning theory and cryptography. In the standard version, we have access to an oracle that, each time we press a button, returns a random vector \(a \in \text{GF}(2)^n\) together with a bit \(b \in \text{GF}(2)\) that was computed as \(a\cdot u + \eta \), where \({u}\in \text{GF}(2)^n\) is a secret vector, and \(\eta \in \text{GF}(2)\) is a noise bit that is 1 with some probability \(p\). Say \(p = 1/3\). The goal is to recover \(u\). This task is conjectured to be intractable.
In the structured noise setting we introduce a slight (?) variation of the model: upon pressing a button, we receive (say) 10 random vectors \({a_1}, {a_2}, \ldots, {a_{10}} \in \text{GF}(2)^n\), and corresponding bits \(b _{1}, b _{2}, \cdots , b _{10}\), of which at most 3 are noisy. The oracle may arbitrarily decide which of the 10 bits to make noisy. We exhibit a polynomial-time algorithm to recover the secret vector \(u\) given such an oracle. We think this structured noise model may be of independent interest in machine learning.
We discuss generalizations of our result, including learning with more general noise patterns. We also give the first nontrivial algorithms for two problems, which we show fit in our structured noise framework.
We give a slightly subexponential algorithm for the well-known learning with errors (LWE) problem over \(\text{GF}(q)\) introduced by \textit{O. Regev} [J. ACM 56, No. 6, Article No. 34, 40 p. (2009; Zbl 1325.68101)] for cryptographic uses. Our algorithm works for the case when the Gaussian noise is small; which was an open problem. Our result also clarifies why existing hardness results fail at this particular noise rate.
We also give polynomial-time algorithms for learning the MAJORITY OF PARITIES function of \textit{B. Applebaum} et al. [in: Proceedings of the 42nd annual ACM symposium on theory of computing, STOC '10. New York, NY: Association for Computing Machinery (ACM). 171--180 (2010; Zbl 1293.94052)] for certain parameter values. This function is a special case of Goldreich's pseudorandom generator.
The full version is available at \url{http://www.eccc.uni-trier.de/report/2010/066/}.
For the entire collection see [Zbl 1217.68003].
Zbl 1325.68101; Zbl 1293.94052