The efficient treatment of strong correlation in molecules is the foremost challenge in the accurate simulation of chemical systems at the quantum level. One of the most efficient methods for treating strong correlation is the “complete active space self-consistent field” (CASSCF) method, which places priors on the Hilbert space of electron configurations in which to carry out the calculation through use of an “active space” of electrons an orbitals in which to diagonalize the Hamiltonian. The problem of efficiently and automatically determining this active space, known as the “active space selection problem”, is the focus of this thesis. Chapters 2, 3, and 4 principally concern the development of automated methods for efficiently determining the active space, while chapters 5, 6, and 7 concern the applications of these approaches.