Memory-Sample Lower Bounds for Learning Parity with Noise

Abstract: In this work, we show, for the well-studied problem of learning parity under noise, where a learner tries to learn $x=(x_1,\ldots,x_n) \in {0,1}<sup>n$ from a stream of random linear equations over $\mathrm{F}_2$ that are correct with probability $\frac{1}{2}+\varepsilon$ and flipped with probability $\frac{1}{2}-\varepsilon$, that any learning algorithm requires either a memory of size $\Omega(n<sup>2/\varepsilon)$ or an exponential number of samples. In fact, we study memory-sample lower bounds for a large class of learning problems, as characterized by [GRT'18], when the samples are noisy. A matrix $M: A \times X \rightarrow {-1,1}$ corresponds to the following learning problem with error parameter $\varepsilon$: an unknown element $x \in X$ is chosen uniformly at random. A learner tries to learn $x$ from a stream of samples, $(a_1, b_1), (a_2, b_2) \ldots$, where for every $i$, $a_i \in A$ is chosen uniformly at random and $b_i = M(a_i,x)$ with probability $1/2+\varepsilon$ and $b_i = -M(a_i,x)$ with probability $1/2-\varepsilon$ ($0&lt;\varepsilon&lt; \frac{1}{2}$). Assume that $k,\ell, r$ are such that any submatrix of $M$ of at least $2^{-k} \cdot |A|$ rows and at least $2^{-\ell} \cdot |X|$ columns, has a bias of at most $2<sup>{-r}$. We show that any learning algorithm for the learning problem corresponding to $M$, with error, requires either a memory of size at least $$\Omega\left(\frac{k \cdot \ell}{\varepsilon} \right)$$, or at least $2<sup>{\Omega(r)}$ samples. In particular, this shows that for a large class of learning problems, same as those in [GRT'18], any learning algorithm requires either a memory of size at least $$\Omega\left(\frac{(\log |X|) \cdot (\log |A|)}{\varepsilon}\right)$$ or an exponential number of noisy samples. Our proof is based on adapting the arguments in [Raz'17,GRT'18] to the noisy case.