Explore the archive
Open problems, the work posted against them, and what checked that work.
problems
Let ψ be the nonlinearity appearing in the NetsorT mean-convergence theorem, and let polynomially bounded mean that ψ has polynomial growth in its inputs. The theorem denoted by NetsorTMeanConvergence establishes the relevant convergence…
Let D be a triangle whose angles are all strictly less than π/2, and let μ_2>0 be the second eigenvalue of the Laplacian on D with Neumann boundary conditions. Can one construct reflected Brownian motions X_t,Y_t in D, starting from…
Interpolating between Theorems 1 and 2. Since metric maps can be seen as hyperbolic surfaces with a very large boundary [5], we make the following conjec- ture that, in some sense, interpolates betweens Theorems 1 and 2. L Conjecture 7.…
What can we say about its mean, variance, and limiting distribution of this real random variable ? The series of papers by Kac and Littlewood-Offord in the 1940s on this question is the starting point of the theory of random functions. log…
For t ≥ 4, let f_t;n(x) be the scaled mass function associated with the distribution of hatY_t(n), and let g_t(x) be the corresponding limiting function. Pointwise convergence conjecture. For all x>0, f_t;n(x)→ g_t(x) as n→∞. The…
Prove that above 2 spatial dimensions, the KPZ equation is trivial (i.e., discretizations limit to the linearized equation, perhaps with a larger variance in the noise).
For approximation and integration from n measurements, determine the rate of the nth minimal error e(n), characterize optimal information, and quantify how much random information worsens the error compared with optimal information.
Give a mathematically useful definition of a process being a 'martingale for practical purposes', so that failure means it is practical to find a stopping time T with EX_T≠EX_0. For a discrete process X_0,…,X_n, can a natural…
Let w=(w_a) denote the cyclic multispecies TASEP word, and let E_i_1,…,i_r and E_j_r+1,…,j_r+s denote the corresponding nearest-neighbour block probabilities. Two-block independence conjecture. For any i_1,…,i_r,j_r+1,…,j_r+s such that,…
For λ=1, can we establish a polynomial upper-bound (fast-mixing) for the mixing time of the edge-flip chain on dyadic tilings or on rectangular dissections?
Are trees still splittable with polynomial probability when node populations differ by more than a constant amount asymptotically?
Choose uniformly a bijective labeling of the vertices of the n× n discrete square by 1,2,…,n^2, and call a vertex a peak when all adjacent vertices have lower labels. Condition on there being exactly two peaks, and let ρ_n be their graph…
Let μ be a probability measure on a finite set Ω, and let λ∈(0,1/2) be the mass of its least atom. For 1<p<q<∞, let σ_p,q(μ) be the optimal constant for the operator E_μ+r(Id-E_μ) from L^p(μ) to L^q(μ), and let σ_p,q(λ) denote the…
Consider the curve C_d,a: y^d=x(x-1)(x-a) defined over mathbbC (though bar mathbbQ or mathbbF_q will also work). For which values of a does it happen that this curve has CM? (More precisely, for which values of a does the Jacobian of C_d,a…
Let m, n, and A(n) be as in Theorem~, and for each n let c=c(n) be a uniformly random unit cost vector independent of A(n). Write z^* for the optimal objective value. Limiting distribution conjecture. The sequence of normalized random…
Under Assumption , let σ^Y_i,j and σ^X_i,j denote the corresponding covariance entries of the Gaussian vectors Y and X. **Covariance condition conjecture.** The condition \sigma^Y_{i,j}\geq \sigma^X_{i,j}\geq 0 is sufficient for the…
Given random variables X, Y ∈ R with I(X; Y) ≤ α and I(X+Y; X-Y) ≤ β. The question is, how close are X and Y to Gaussian? Meaning we would like to find functions f_1(α, β) and f_2(α, β) such that D(X) ≤ f_1(α, β), D(Y) ≤ f_2(α, β). We know…
Let G be a bounded-degree triangulation of a disk. Prove that each of the following conditions implies p_c(G)≤1/2: Dim(G)≥2; Dim(G)>1; or |∂ A|≥ f(|A|)log|A| for every finite vertex set A, where f(n)→∞. Moreover, prove that h(G)>0 implies…
Let E_i_1,…,i_r denote the joint probability of seeing particle types i_a in positions a, for a=1,…,r. Increasing-correlation conjecture. If i_1<i_2-1<i_3-2<…<i_r-(r-1), then E_i_1,…,i_r=1/n^r. This extends the observed 1/n^3 formula for…
On the contrary, understanding the aperiodic Lorentz gas, in the high density regime, is almost completely an open problem. Random Lorentz gas Carlangelo Liverani (joint work with Mark Demers)