Sommersemester 2018

Combinatorial optimization problems appear in various application areas, e.g., in logistics, project planning, communication and financial markets. When solving real‐world optimization problems, uncertainty in the input data is a prevalent issue: resources may become unavailable, material arrives late, jobs may take more or less time than originally estimated, etc. In this talk, we will discuss models, such as online, stochastic and robust models, and solution methods for optimization under uncertainty. What is an algorithm that "performs well", even under uncertain input? How can optimization be incorporated in a decision-support system? Exemplarily, we will consider scheduling problems and discuss theoretical performance guarantees and applications in practice.

In everyday life as well as in scientific practice, we tend to take assertions like the following to be at least possibly true: The cause of her death was the failure of her heart; my exercising caused my muscles to become stronger; the alcohol in his blood caused him to stumble and fall. These assertions make claims about interlevel causation. Philosophers are worried about claims like these because, so it is argued, their truth commits us to mysterious causal interactions between wholes and their parts. Furthermore, they are in conflict with the famous Causal Exclusion Argument. This argument shows that higher-level phenomena cannot do any causal work—neither with respect to what is going on at the same level, nor what is going on at lower-levels. Causation takes place only at the fundamental level, rendering all other phenomena either non-existent or epiphenomal. If these philosophers are right, not only would we speak falsely every time we make claims like the ones stated above. Also, the autonomy of higher-level sciences, like chemistry and biology, from fundamental sciences, like particle physics, would be in severe danger. But are these philosophers really right?

In this talk, I will explain first what philosophers mean by “level”, “causation”, and “interlevel causation”. Then, I will introduce the two main objections against interlevel causation. Finally, I will present my theory of interlevel causation which provides solutions to the two objections.

Collective decisions are produced by voting procedures which aggregate individual preferences and judgments. Before and after, individual preferences and judgments change as their underlying attitudes, values, and opinions change through discussion and deliberation. In large groups, these dynamics naturally go beyond the scope of the individual and consequently might show unexpected self-driven macroscopic systems dynamics following “socio-physical” laws.

We model the dynamics of opinion formation through social interaction and consider attitudes on a one dimensional scale which represents for example emotional valence towards something, political left-right selfplacement, or adherence to certain social norms. Individual opinion change is modeled with respect to a reinforcement theory, an information processing theory, a social judgment theory, and a polarity effect. We will answer the questions to what extend these indivual mechanisms contribute to the societal phenomena of extremal consensus, moderate consensus, polarization, fractionalization, maintenance of diversity, and cyclic behavior. The dynamics of the mean and median opinion finally informs us on the implications of these for collective decisions.

Living cells are decision-making units resting on a complex molecular machinery. The machinery has evolved to perform robust biological functionality at the population level or in a full organism. Yet the composition of the molecular network exhibits cell-to-cell variability and is poised by stochastic fluctuations leaving the outcome of an individual cell decision partially erratic. In my talk, I aim at inferring rules of molecular decision making from monitoring the dynamical response of many individual cells to external stimuli under standardized conditions and with ample statistics. Various examples of single cell time-lapse studies are presented that employ micro-structured substrates and fluorescent reporter signals showing genetic switches, apoptotic signaling cascades and cell locomotion. I will conceptualize the notion of cell states and systems biology modeling of functional modules and discuss the implications of probabilistic strategies on our view of decision making.

Der Vortrag entwickelt zunächst eine Perspektive, die Entscheiden als ein prozesshaftes soziales Geschehen betrachtet, das sich auf das Entwickeln von alternativen Handlungsoptionen zu einem bestimmten Thema, auf deren Bewertung sowie die Selektion einer von ihnen bezieht. Kulturen des Entscheidens bringen Entscheiden hervor; sie schließen Sprechakte und Erzählungen mit ein und weisen sowohl symbolisch-expressive als auch technisch-instrumentelle Aspekte auf. Kulturen des Entscheidens variieren nach dem Ort, an dem entschieden wird (Universität, Parlamente, Haushalt), und verändern sich über die Zeit hinweg. Ein zweiter Teil des Vortrags identifiziert deshalb Eigenschaften von Kulturen des Entscheidens, die in vormodernen Kulturen - in denen Entscheiden als Handlungsform weniger leicht verfügbar war - weit verbreitet waren. Gegenstand des dritten und letzten Teils des Vortrags sind schließlich Vorgänge, die langfristig eine moderne Entscheidungsgesellschaft hervorgebracht haben.

ow-driven amoeboid motility is a fundamental process in biology, and it provides and interesting paradigm for bio-inspired design because this motility mode is fast and robust to changes in the extracellular environment. This talk will examine the spatiotemporal coordination of traction stress and intracellular flow driving the amoeboid locomotion of Physarum polycephalum microplasmodia. The cytoplasm of physarum is a mechano-chemically active, complex fluid. Cytoplasmic flow and transport of signaling molecules (e.g. calcium ions) occur in a complex regime affected by both convection and diffusion, and take place on a time-dependent geometry caused by fluid-structure interactions at the cell length-scale. The active stresses in this fluid are regulated by calcium ions, whose propagation is in turn governed by the flow. We study this system by a combination of quantitative live-cell experiments and computational modeling to provide a complete characterization of the mechanics of amoeboid migration with high resolution (~ 5μm in space, ~ 1sec in time), and to explore fundamental design principles that could be extrapolated from this system.

Bio. Prof. del Alamo received a B.S., M.S. and Ph. D. in Aerospace Engineering at the Polytechnic University in Madrid. He was a Fulbright postdoctoral fellow at Harvard University and UC Sand Diego, where he received training in experimental cell mechanics and cardiovascular flows. Prof. del Alamo’s lab at UCSD focuses on biological fluid mechanics, cellular locomotion and non-invasive characterization of cardiac flows.