| In this master's thesis, we are interested in the media industry, particularly in media broadcasting and the associated pricing problem.;For the first time, this revenue management problem was solved thanks to bilevel programming. In this model (called "M-Spot"), the TV channel offers and sells commercial spots to clients. Even if bilevel programming is very convenient as a paradigm for this problem, the model fails to tackle large-scale instances and proposes unrealistic solutions.;Using this model, we present another bilevel program, based on patterns of commercial spots, which we call "M-Pattern". As in the cutting stock problem, we use patterns to cut media inventories. This way, a pattern can be considered as a set of commercials, and TV channels sell selection of commercial spots, whereas in the M-Spot they sell single commercials.;We take up an algorithm developed for the M-Spot model, based on cut generation. We adapt this algorithm to our model, and introduce additional cuts in a pre-processive step. Both algorithms allow to optimally schedule and price patterns of commercial spots for a typical prime time on TV.;TV Channels face many problems when they sell their inventory to media agencies. On one hand, they have to choose which commercials should be aired. On the other hand, they must determine the prices of these commercials, while staying competitive.;Keywords: Advertising, revenue management, bilevel programming, commercial, pattern, knapsack, branch-and-cut. |
