Investigating solutions of nonlinear equation systems is challenging in a general framework, especially if the equations contain uncertainties about parameters modeled by probability densities. Such random equations, understood as stationary (non-dynamical) equations with parameters as random variables, have a long history and a broad range of applications. In this work, we study nonlinear random equations by combining them with mixture model parameter random variables in order to investigate the combinatorial complexity of such equations and how this can be utilized practically. We derive a general likelihood function and posterior density of approximate best fit solutions while avoiding significant restrictions about the type of nonlinearity or mixture models, and demonstrate their numerically efficient application for the applied researcher. In the results section we are specifically focusing on example simulations of approximate likelihood/posterior solutions for random linear equation systems, nonlinear systems of random conic section equations, as well as applications to portfolio optimization, stochastic control and random matrix theory in order to show the wide applicability of the presented methodology.
翻译:暂无翻译