Development of optimization methodology for modeling large-scale problems in decision or control over time, possibly with stochastic elements; numerical techniques for solving such problems; and associated innovations in mathematical analysis. Current problems include: extensions of linear and quadratic programming to provide more robust treatment of problem models in multistage optimization and optimal control; stochastic programming (a fast-developing subject which concerns situations where corrective decisions can be taken at various future times in response to observations of uncertain phenomena); and applications of stochastic programming to finance and economics. ┬áProfessor Rockafellar’s research sponsors are the National Science Foundation and the Air Force Office of Scientific Research.