Search

We introduce Reactive Message Passing (RMP) as a framework for executing schedule-free, robust and scalable message passing-based inference in a factor graph representation of a probabilistic model. RMP is based on the reactive programming style that only describes how nodes in a factor graph react to changes in connected nodes. The absence of a fixed message passing schedule improves robustness, scalability and execution time of the inference procedure. We also present ReactiveMP.jl, which is a Julia package for realizing RMP through minimization of a constrained Bethe free energy. By user-defined specification of local form and factorization constraints on the variational posterior distribution, ReactiveMP.jl executes hybrid message passing algorithms including belief propagation, variational message passing, expectation propagation, and expectation maximisation update rules. Experimental results demonstrate the improved performance of ReactiveMP-based RMP in comparison to other Julia packages for Bayesian inference across a range of probabilistic models. In particular, we show that the RMP framework is able to run Bayesian inference for large-scale probabilistic state space models with hundreds of thousands of random variables on a standard laptop computer.

Gesture recognition enables a natural extension of the way we currently interact with devices. Commercially available gesture recognition systems are usually pre-trained and offer no option for customization by the user. In order to improve the user experience, it is desirable to allow end users to define their own gestures. This scenario requires learning from just a few training examples if we want to impose only a light training load on the user. To this end, we propose a gesture classifier based on a hierarchical probabilistic modeling approach. In this framework, high-level features that are shared among different gestures can be extracted from a large labeled data set, yielding a prior distribution for gestures. When learning new types of gestures, the learned shared prior reduces the number of required training examples for individual gestures. We implemented the proposed gesture classifier for a Myo sensor bracelet and show favorable results for the tested system on a database of 17 different gesture types. Furthermore, we propose and implement two methods to incorporate the gesture classifier in a real-time gesture recognition system.

The economic crisis that started in 2007 draws new attention to the financialization of the economy and in particular to the concept of debt. Simultaneously, rising energy and commodity prices are yet another indication for concern about a growing population, emissions and resource exploitation that is burdening the Earth. This thesis provides an extensive overview and analysis of the relevant literature, followed by the application of economist Steve Keen’s model of the role of money in a simple credit economy. It is used to test whether the money system is viable and/or has inherent mechanisms that hinder sustainable development. Both methods aim to be a springboard from which further research can be undertaken as this remains a topic alive with uncertainty. I find that although the money system does not mechanically prevent sustainable development, it does enable and promote a tendency towards inequality, a growth imperative, and an economically unsustainable debt burden. It thus does hinder sustainable development.

Never before in history was the interaction between people and their natural environment as complex and problematic as it is today. A proliferation of scientific research has yielded valuable insights into various aspects of this interaction from the angle of many disciplines - the natural sciences, the social sciences, archaeology and history, ecological studies. The diversity of approaches has created a need for synthesis, for a study that transcends the boundaries of traditional fields of study. In this volume, authors from various academic backgrounds discuss the relations between human society and its physical environment in the course of history, highlighting a number of significant periods, throughout the world. The last chapter assesses our present situation and prospects for the future in the light of theoretical reflections based on the evidence from the past.

PR, IFPRI4; Theme 1; Subtheme 1.1; Global food scenarios, EPTD

Nooit eerder in de geschiedenis was de interactie tussen de mens en zijn natuurlijke omgeving zo complex en problematisch als vandaag de dag. De explosieve groei van wetenschappelijk onderzoek heeft waardevolle inzichten opgeleverd over de meest uiteenlopende aspecten van deze wisselwerking: zo zijn de mens en de natuur bestudeerd vanuit de optiek van de natuurwetenschappen, de sociale wetenschappen, archeologie, geschiedenis en ecologie. Door deze enorme verscheidenheid aan inzichten is er een een behoefte aan synthese ontstaan, aan een onderzoek dat buiten de traditionele gebaande studie-paden gaat. In dit boek wordt door auteurs met verschillende wetenschappelijke achtergronden een discussie gevoerd over de relatie tussen de menselijke samenleving en zijn fysieke omgeving. Vanuit historisch perspectief wordt een aantal veelbetekendende periodes over de gehele wereld besproken. Het laatste hoofdstuk gaat in op onze huidige situatie en de vooruitzichten voor de toekomst aan de hand van theoretische reflecties gebaseerd op de lessen van het verleden