Résumé
The third lecture will discuss the Erdös-Ginzburg-Ziv Problem in discrete geometry, and show how (at least in high dimension) it can be reformulated as a problem in F_p^n. Specifically, this leads to the question of asking about the maximum possible size of a subset of F_p^n not containing p distinct vectors summing to zero. Some bounds for this problem will be deduced from the Ellenberg-Gijswijt result for three-term progression free sets in F_p^n discussed in the previous lecture.
