Optimal control in linear stochastic advertising models with memory

TL;DR

This paper develops a novel infinite-dimensional approach to solve optimal control problems in stochastic advertising models with memory effects, using kernel approximation techniques to handle complex Volterra processes.

Contribution

It introduces a method to address optimal control in stochastic models with memory by approximating kernels and formulating an equivalent Hilbert space problem.

Findings

The method effectively solves control problems with H"older continuous kernels.

Kernel approximation via Bernstein polynomials simplifies the control problem.

Simulations demonstrate the practical applicability of the approach.

Abstract

This paper deals with a class of optimal control problems which arises in advertising models with Volterra Ornstein-Uhlenbeck process representing the product goodwill. Such choice of the model can be regarded as a stochastic modification of the classical Nerlove-Arrow model that allows to incorporate both presence of uncertainty and empirically observed memory effects such as carryover or distributed forgetting. We present an approach to solve such optimal control problems based on an infinite dimensional lift which allows us to recover Markov properties by formulating a optimization problem equivalent to the original one in a Hilbert space. Such technique, however, requires the Volterra kernel from the forward equation to have a representation of a particular form that may be challenging to obtain in practice. We overcome this issue for H\"older continuous kernels by approximating them with Bernstein polynomials (which turn out to enjoy a simple representation of the required type) and then solving the optimal control problem for the forward process with approximated kernel instead of the original one. The approach is illustrated with simulations.