Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
The authors build dimensionally homogeneous surrogate models for physical systems by hardcoding unit-invariance directly into the hypothesis class. By combining Buckingham $\Pi$-theorem reductions with admissible dimensional prefactors, the remaining dimensionless dynamics are mapped onto normalized domains and approximated via truncated harmonic expansions. Evaluated on systems ranging from Planck’s black-body radiation to double-pendulum Lyapunov fields, the constrained formulation yields strictly superior conditioning, sample efficiency, and noise robustness compared to unconstrained baselines.
Baking dimensional homogeneity directly into surrogate hypothesis spaces turns complex physical system identification into a robust, highly sample-efficient linear regression problem.
Dimensional homogeneity is a fundamental constraint on physically meaningful models, requiring invariance under changes of units. We present a data-driven method for constructing surrogate models that satisfy this constraint at the level of the hypothesis class. Starting from a dimension matrix of measured variables, the method derives Buckingham $Π$-groups, constructs admissible dimensional prefactors, and approximates the remaining dimensionless dependence using truncated harmonic expansions on normalized invariant domains. Once the prefactor and dictionary are fixed, the coefficients are obtained from a regularized linear regression problem. We test the approach on the simple pendulum, Planck's black-body law, the double-pendulum Lyapunov field, and an experimental COBE/FIRAS black-body spectrum dataset. The results show that dimensional constraints improve conditioning, robustness to noise, and sample efficiency relative to unconstrained baselines, while the choice of dictionary becomes important in non-periodic or multi-invariant settings. The learned expressions are explicit and inexpensive to evaluate, which makes them useful as surrogate models for structured physical problems.