Track new results across pure and applied mathematics.
Search papers, labs, and topics across Lattice.
This paper introduces a novel framework for Gaussian Process (GP) solvers of linear PDEs that incorporates linear inequality constraints, moving beyond traditional equality-only conditioning. By proving an equivalence between joint and sequential conditioning on equality and inequality events, the authors enable the integration of constraints like nonnegativity and monotonicity directly into the GP posterior. This approach, computable via truncated-Gaussian moments, significantly improves accuracy and adherence to physical laws, as demonstrated on benchmarks for Poisson equations, option pricing, and air-quality modeling.
Inequality constraints, previously a post-processing step, now directly inform probabilistic PDE solvers, slashing reconstruction errors by up to 63% and arbitrage rates by 90%.
Gaussian process solvers for linear partial differential equations (PDEs) return posteriors rather than point estimates, but they condition only on equality information: the PDE residual, boundary data, and measurements. Many solutions of interest also obey inequality constraints, including nonnegativity, monotonicity, and absence of arbitrage, which equality conditioning cannot enforce. We prove that conditioning a Gaussian process prior jointly on linear equality observations and a finite-dimensional inequality event is equivalent to conditioning sequentially in terms of Borel measures on the solution space. The physics-informed posterior is therefore a valid prior for the inequality update at virtual points. The combined law is non-Gaussian yet computable through truncated-Gaussian moments, and neither the PDE conditioning nor the inequality update requires modification. We evaluated the construction on a Poisson benchmark, option-price surfaces, and air-quality fields. Under incomplete boundary data, a monotonicity constraint lowered the error of a two-dimensional Poisson reconstruction by $63\%$. On a real S&P 500 option chain thinned to $20$ retained quotes, no-arbitrage constraints reduced held-out extrapolation error by $22\%$ and cut the calendar-arbitrage rate from $10\%$ to $1\%$. On four real air-quality snapshots, nonnegativity reduced a Gaussian moment-based sub-zero-tail proxy by $74\%$ at the sparsest monitor count. Changes in mean accuracy and constraint diagnostics differed across data regimes. The probability of the noisy virtual constraint event under the equality posterior measures agreement with the specified bounds before the update. The construction turns admissibility from a post-processing step into conditioning information for probabilistic PDE solvers.