← Back to Problems
Probability TheoryResearchAI-Generated

Does a finite-moment criterion analogous to the four-moment theorem for Gaussian chaos characterize Poisson convergence on general discrete chaoses beyond the Poisson and Rademacher settings?

Related: Nualart-Peccati fourth-moment theorem, Stein-Chen method for Poisson approximation, de Jong central limit theorem for degenerate U-statistics

Resolving this question would establish a universal finite-moment principle for chaos convergence across the full landscape of discrete probability, mirroring the completeness that the fourth-moment theorem brought to Gaussian analysis. It would create practical statistical tests for Poisson-type behavior in systems built from non-standard noise, relevant to random graph theory, combinatorial stochastic processes, and high-dimensional statistics. It would also clarify the deeper reason why four moments suffice at all, potentially revealing a unifying algebraic obstruction tied to the genus of interaction diagrams rather than to the specific choice of noise.

View Source Paper →