Optimization in Online Advertising via Simultaneous Adaptive Rate and Price Feedback Control

Online advertising is typically implemented via real-time bidding, and advertising campaigns are then defined as extremely high-dimensional optimization problems. To solve these problems in light of large scale and significant uncertainties, the optimization problems are modularized in a way that makes feedback control a critical component of the solution. The control problem, however, is challenging due to plant uncertainties, nonlinearities, time-variance, and noise. An Oracle would define the control signal in terms of bid price adjustments only; however, we propose the introduction of a companion throttling control signal that creates a useful plant linearity. In this paper, the control problem is redefined in such a way that the linearity is exploited for improved feedback control. A dual lever control algorithm is designed and evaluated in simulations, with promising results.