Timezone: »

On the Cryptographic Hardness of Learning Single Periodic Neurons
Min Jae Song · Ilias Zadik · Joan Bruna

Tue Dec 07 04:30 PM -- 06:00 PM (PST) @ None #None
We show a simple reduction which demonstrates the cryptographic hardness of learning a single periodic neuron over isotropic Gaussian distributions in the presence of noise. More precisely, our reduction shows that any polynomial-time algorithm (not necessarily gradient-based) for learning such functions under small noise implies a polynomial-time quantum algorithm for solving worst-case lattice problems, whose hardness form the foundation of lattice-based cryptography. Our core hard family of functions, which are well-approximated by one-layer neural networks, take the general form of a univariate periodic function applied to an affine projection of the data. These functions have appeared in previous seminal works which demonstrate their hardness against gradient-based (Shamir'18), and Statistical Query (SQ) algorithms (Song et al.'17). We show that if (polynomially) small noise is added to the labels, the intractability of learning these functions applies to all polynomial-time algorithms, beyond gradient-based and SQ algorithms, under the aforementioned cryptographic assumptions. Moreover, we demonstrate the necessity of noise in the hardness result by designing a polynomial-time algorithm for learning certain families of such functions under exponentially small adversarial noise. Our proposed algorithm is not a gradient-based or an SQ algorithm, but is rather based on the celebrated Lenstra-Lenstra-Lov\'asz (LLL) lattice basis reduction algorithm. Furthermore, in the absence of noise, this algorithm can be directly applied to solve CLWE detection (Bruna et al.'21) and phase retrieval with an optimal sample complexity of $d+1$ samples. In the former case, this improves upon the quadratic-in-$d$ sample complexity required in (Bruna et al.'21).

Author Information

Min Jae Song (New York University)
Ilias Zadik (New York University)

I am a CDS Moore-Sloan (postdoctoral) fellow at the Center for Data Science of NYU and a member of it's Math and Data (MaD) group. I received my PhD on September 2019 from MIT , where I was advised by David Gamarnik. My research lies broadly in the interface of high dimensional statistics, the theory of machine learning and applied probability.

Joan Bruna (NYU)

More from the Same Authors