On Learning Parallel Pancakes with Mostly Uniform Weights

We study the complexity of learning $k$-mixtures of Gaussians ($k$-GMMs) on $\mathbb{R}^d$. This task is known to have complexity $d^{\Omega(k)}$ in full generality. To circumvent this exponential lower bound on the number of components, research has focused on learning families of GMMs satisfying additional structural properties. A natural assumption posits that the component weights are not exponentially small and that the components have the same unknown covariance. Recent work gave a $d^{O(\log(1/w_{\min}))}$-time algorithm for this class of GMMs, where $w_{\min}$ is the minimum weight. Our first main result is a Statistical Query (SQ) lower bound showing that this quasi-polynomial upper bound is essentially best possible, even for the special case of uniform weights. Specifically, we show that it is SQ-hard to distinguish between such a mixture and the standard Gaussian. We further explore how the distribution of weights affects the complexity of this task. Our second main result is a quasi-polynomial upper bound for the aforementioned testing task when most of the weights are uniform while a small fraction of the weights are potentially arbitrary.

@article{diakonikolas2025_2504.15251,
  title={ On Learning Parallel Pancakes with Mostly Uniform Weights },
  author={ Ilias Diakonikolas and Daniel M. Kane and Sushrut Karmalkar and Jasper C.H. Lee and Thanasis Pittas },
  journal={arXiv preprint arXiv:2504.15251},
  year={ 2025 }
}