Game theoretical methods in PDEs (Q2257090)

The authors describe how the solutions to certain PDEs of \(p\)-Laplacian type can be interpreted as limits of values of a specific Tug-of-War game, when the step-size \(\epsilon\) determining the allowed length of move of a token, decreases to 0. After explaining in Section 1, how linear elliptic equations arise in probability, they describe in Section 2 how \(p\)-harmonic functions can be approximated by functions known as \(p\)-harmonious functions defined by the mean value property. These \(p\)-harmonious functions have a probabilistic interpretation as values of Tug-of-War games with noise explained in Section 3. Using the observation that a sequence of random variables involving \(p\)-harmonious functions is supermartingale, they argue that the minimum gain of Player I and the maximimum loss of Player II in a Tug-of-War game with noise is equal to the \(p\)-harmonious function. In Section 5 the authors discuss the example of the proof of uniform convergence and the local Harnack inequality using the technique of assigning suitable strategies in a Tug-of-War game.