The management of food supply chains is recently receiving increased attention, both in practice and in scientific literature. Food supply chains show specific characteristics that lead to challenging research questions. Food products show continuous quality changes throughout the supply chain, all the way until final consumption by human beings. Hence, quality, health, and safety require central consideration. The production and distribution of food in the current globalized economy is a major discussion point in society. The food industry has been at the forefront in developments related to sustainability (reflected, for example, in the discussions about food miles). This talk will introduce optimization modelling concepts designed to address the challenges of food supply chain management.
Dr. Jannik Matuschke: "A Local Search Algorithm for Steiner Forest"
Local search is a widely used concept for solving optimization problems. Local search algorithms gradually improve the present solution by applying small ('local') changes. Unfortunately, it is well-known that a local optimum found in this way can be significantly inferior to the global optimum of the problem. We present a new local search algorithm for Steiner Forest, a fundamental network design problem that asks for connecting a collection of terminal pairs at minimum cost. We show how bad local optima can be avoided by using an additional penalty term in the objective, obtaining a constant factor approximation for the problem.
This is joint work with Martin Groß, Anupam Gupta, Amit Kumar, Daniel R. Schmidt, Melanie Schmidt, and José Verschae.
Venue: TUM Campus Innenstadt Room Z536 (Arcisstr. 21 [505], 1st intermediate storey)
Date: Monday, February 5th, 2018, 14:00