1: \begin{abstract}

2: A non-linear conjugate gradient optimization scheme is used to obtain excitation energies within the Random Phase Approximation (RPA).

3: The solutions to the RPA eigenvalue equation are located through a variational characterization using a modified Thouless functional, which is based upon an asymmetric Rayleigh quotient, in an orthogonalized atomic orbital representation.

4: In this way, the computational bottleneck of calculating molecular orbitals is avoided.

5: The variational space is reduced to the physically-relevant transitions by projections.

6: The feasibility of an RPA implementation scaling linearly with system size, $N$, is investigated by monitoring  convergence behavior with respect to the quality of initial guess and sensitivity to noise under thresholding, both for well- and ill-conditioned problems.

7: The molecular-orbital-free algorithm is found to be robust and computationally efficient providing a first step toward a large-scale, reduced complexity calculation of time-dependent optical properties and linear response. 

8: The algorithm is extensible to other forms of time-dependent perturbation theory including, but not limited to, time-dependent Density Functional theory.

9: 

10: \end{abstract}

11: