The optimal Jacobian accumulation problem is one of the original
combinatorial problems arising in automatic differentiation.
The vast majority of approaches to this problem have attempted to
exploit its similarity to the problem of minimizing fill during LU
factorization of sparse, unsymmetric matrices.
In this talk, we take a complexity-theoretic approach to Jacobian accumulation.
Specifically, we explore relationships between this problem and others
that occur in the context of algebraic and Boolean complexity,
as well as discuss the results that are implied by these relationships.