1,090 research outputs found
Author Interview with Brian Klaas: how Can We Fix Democracy?
Full text linkThe end of the Cold War in the early 1990s saw democracy surge as former Soviet autocracies transitioned to democratic systems and democracy spread in Africa and Latin America. But the past decade has seen a reverse in this trend, with authoritarianism and dictatorships making a comeback around the world. In this interview with Peter Carrol on his new book, The Despot’s Accomplice: How the West is Aiding and Abetting the Decline of Democracy, Dr Brian Klaas (LSE Department of Government) argues that Western powers are partly to blame for these developments, and offers a number of solutions to halt the decline of democracy around the world
A Bayesian foundation for individual learning under uncertainty
Full text linkComputational learning models are critical for understanding mechanisms of adaptive behavior. However, the two major current frameworks, reinforcement learning (RL) and Bayesian learning, both have certain limitations. For example, many Bayesian models are agnostic of inter-individual variability and involve complicated integrals, making online learning difficult. Here, we introduce a generic hierarchical Bayesian framework for individual learning under multiple forms of uncertainty (e.g., environmental volatility and perceptual uncertainty). The model assumes Gaussian random walks of states at all but the first level, with the step size determined by the next higher level. The coupling between levels is controlled by parameters that shape the influence of uncertainty on learning in a subject-specific fashion. Using variational Bayes under a mean field approximation and a novel approximation to the posterior energy function, we derive trial-by-trial update equations which (i) are analytical and extremely efficient, enabling real-time learning, (ii) have a natural interpretation in terms of RL, and (iii) contain parameters representing processes which play a key role in current theories of learning, e.g., precision-weighting of prediction error. These parameters allow for the expression of individual differences in learning and may relate to specific neuromodulatory mechanisms in the brain. Our model is very general: it can deal with both discrete and continuous states and equally accounts for deterministic and probabilistic relations between environmental events and perceptual states (i.e., situations with and without perceptual uncertainty). These properties are illustrated by simulations and analyses of empirical time series. Overall, our framework provides a novel foundation for understanding normal and pathological learning that contextualizes RL within a generic Bayesian scheme and thus connects it to principles of optimality from probability theory
Monitoring outpatients in palliative care through wearable devices
Full text linkPatients in palliative care suffer from a life-threatening disease. Holistic treatment includes control of symptoms (e. g., pain, nausea, sleeplessness) as well as psychosocial and spiritual help which is also extended to the relatives of a patient. For advanced cancer patients in palliative care, a crucial phase is the transition from palliative care in the hospital to the home setting, where care around the clock is not guaranteed any more, leads to an increased number of unplanned hospital re-admissions and emergency visits. Physicians aim to fill this care gap by monitoring physical and social activities as well as vital signs. Daily monitoring data, provided to caregivers, could enable caregivers to timely intervene when symptoms of a patient deteriorate.
Besides patients in palliative care, also cancer survivors suffering from cancer-related fatigue could benefit from activity monitoring. Up to now, the remedies and effective treatments for cancer-related fatigue are limited. Research still has to unveil the underlying mechanisms that lead to a state of chronic exhaustedness. Measures that help healthy people like regenerative sleep show no or little effect in fatigued patients. Besides psycho-stimulants that come with the risk of addiction, cognitive behavioural therapy and moderate physical exercise have been shown to be effective. However, research still has to investigate timing, frequency and intensity of physical activity and researchers need a better understanding how the fatigue evolves during the day and in long-term.
This thesis investigates the possibilities and limitations of activity monitoring using wearable devices such as smartphones and an armworn devices that is capable of measuring vital signs such as heart rate. Three studies involving cancer patients are conducted:
- An interview study including 12 cancer patients enabled a patient-centric design for an Android activity monitoring app for smartphones.
- Only using the smartphone as monitoring device, a study with 7 cancer survivors suffering from cancer-related fatigue was conducted as a pre-study in order to gain first experiences and to explore the possible knowledge gain about cancer-related fatigue through activity monitoring.
- During a planned study period of 12 weeks per patient, 30 patients in ambulatory palliative care were wearing a smartphone and the arm-worn sensor as monitoring devices. The age range of the study participants was 39 to 85 years. In weekly interviews, patients were asked about their experiences with the devices and their quality of life. The aim of the study was to evaluate feasibility and acceptance of activity monitoring in this patient group.
Furthermore, exploratory data analysis investigated the possibilities and limitations of unsupervised methods on this real-world data set. The two data sets, collected during the fatigue study and during the palliative care study, were pre-processed including cleaning steps, classification and clustering methods to add higher level information such as visited locations (anonymized). From these prepared data sets, features were extracted such as number of places visited per day. On the resulting datasets of features, statistical methods were applied to explore relations between sensor data, self-reports and, in case of the palliative care study, emergency visits to the hospital. For the latter analysis, patients who experienced an emergency room visit and those who did not were compared by means of hypothesis testing. For each feature, the underlying alternative hypothesis was that the change of a feature between the first week of study participation at home and the week before an emergency visit (or the last week of study participation for the patients without an emergency visit), differs in the two patient groups. The rate of change was defined by the ratio of the medians of the two weeks.
Changes of three features, namely resting heart rate, resting heart rate variability and step speed were identified to have significant group differences:
- The resting heart rate had an increasing trend in the group with emergency visits (median=1.01, interquartile range [0.96, 1.12]) and a decreasing trend in the group without an emergency visit (median=0.9, interquartile range [0.89, 0.99]) with a nominal significance of p=.021 and a medium effect size r=.46.
- The resting heart rate variability had a decreasing trend in the group with emergency visits (mean=0.81, standard deviation=0.14) and an increasing trend in the group without an emergency visit (mean=1.17, standard deviation=0.46) with a nominal significance of p=.011 and a large effect size r=.53.
- The step speed had an increasing trend in the group with emergency visits (median=1.1, interquartile range [1.08, 1.13]) and a decreasing trend in the group without an emergency visit (median=0.99, interquartile range [0.96, 1.04]) with a nominal significance of p=.003 and a large effect size r=.61.
In contrast, hypothesis testing for features based on patients’ subjective self-reports for pain, distres and global quality of life did not reveil any significant differences. Hence, activity monitoring of vital signs and physical activity outperformed patients’ self-reports. However, a power analysis based on the three nominally significant results would recommend an independent study with 84 patients to confirm the results of this study.
Furthermore, a set of recommendations for future research was concluded from the experiences gained through conducting these studies
A Large-Scale 3D Study on Transport of Humic Acid-Coated Goethite Nanoparticles for Aquifer Remediation
Full text linkfirst_page
settings
Open AccessFeature PaperArticle
A Large-Scale 3D Study on Transport of Humic Acid-Coated Goethite Nanoparticles for Aquifer Remediation
by Milica Velimirovic 1,2 [OrcID] , Carlo Bianco 3 [OrcID] , Natalia Ferrantello 3, Tiziana Tosco 3, Alessandro Casasso 3 [OrcID] , Rajandrea Sethi 3,*, Doris Schmid 1, Stephan Wagner 1,4 [OrcID] , Kumiko Miyajima 5, Norbert Klaas 5, Rainer U. Meckenstock 6 [OrcID] , Frank von der Kammer 1 and Thilo Hofmann 1,*
1
Department of Environmental Geosciences, Centre for Microbiology and Environmental Systems Science, University of Vienna, Althanstrasse 14, 1090 Vienna, Austria
2
Department of Chemistry, Atomic & Mass Spectrometry–A&MS Research Group, Campus Sterre, Ghent University, Krijgslaan 281-S12, 9000 Ghent, Belgium
3
Department of Environmental, Land and Infrastructure Engineering (DIATI), Politecnico di Torino, corso Duca degli Abruzzi, 24-10129 Torino, Italy
4
Department of Analytical Chemistry, UFZ-Helmholtz Centre for Environmental Research, Permoserstrasse 15, 04318 Leipzig, Germany
5
VEGAS—Research Facility for Subsurface Remediation, University of Stuttgart, Pfaffenwaldring 61, 70569 Stuttgart, Germany
6
Environmental Microbiology and Biotechnology, University Duisburg-Essen, 45141 Essen, Germany
*
Authors to whom correspondence should be addressed.
Water 2020, 12(4), 1207; https://doi.org/10.3390/w12041207
Received: 17 March 2020 / Revised: 10 April 2020 / Accepted: 16 April 2020 / Published: 24 April 2020
(This article belongs to the Special Issue Groundwater and Soil Remediation)
Download PDF Browse Figures
Review Reports Cite This Paper
Abstract
Humic acid-coated goethite nanoparticles (HA-GoeNPs) have been recently proposed as an effective reagent for the in situ nanoremediation of contaminated aquifers. However, the effective dosage of these particles has been studied only at laboratory scale to date. This study investigates the possibility of using HA-GoeNPs in remediation of real field sites by mimicking the injection and transport of HA-GoeNPs under realistic conditions. To this purpose, a three-dimensional (3D) transport experiment was conducted in a large-scale container representing a heterogeneous unconfined aquifer. Monitoring data, including particle size distribution, total iron (Fetot) content and turbidity measurements, revealed a good subsurface mobility of the HA-GoeNP suspension, especially within the higher permeability zones. A radius of influence of 2 m was achieved, proving that HA-GoeNPs delivery is feasible for aquifer restoration. A flow and transport model of the container was built using the numerical code Micro and Nanoparticle transport Model in 3D geometries (MNM3D) to predict the particle behavior during the experiment. The agreement between modeling and experimental results validated the capability of the model to reproduce the HA-GoeNP transport in a 3D heterogeneous aquifer. Such result confirms MNM3D as a valuable tool to support the design of field-scale applications of goethite-based nanoremediation
Variational Bayesian mixed-effects inference for classification studies
Full text linkMultivariate classification algorithms are powerful tools for predicting cognitive or pathophysiological states from neuroimaging data. Assessing the utility of a classifier in application domains such as cognitive neuroscience, brain–computer interfaces, or clinical diagnostics necessitates inference on classification performance at more than one level, i.e., both in individual subjects and in the population from which these subjects were sampled. Such inference requires models that explicitly account for both fixed-effects (within-subjects) and random-effects (between-subjects) variance components. While models of this sort are standard in mass-univariate analyses of fMRI data, they have not yet received much attention in multivariate classification studies of neuroimaging data, presumably because of the high computational costs they entail. This paper extends a recently developed hierarchical model for mixed-effects inference in multivariate classification studies and introduces an efficient variational Bayes approach to inference. Using both synthetic and empirical fMRI data, we show that this approach is equally simple to use as, yet more powerful than, a conventional t-test on subject-specific sample accuracies, and computationally much more efficient than previous sampling algorithms and permutation tests. Our approach is independent of the type of underlying classifier and thus widely applicable. The present framework may help establish mixed-effects inference as a future standard for classification group analyses
Tomographic particle-image velocimetry measurements in a turbulent wavy channel flow
No full textTo investigate small scale turbulent structures and their statistical properties in non-isotropic turbulent flow subjected to favorable and adverse pressure gradients, a novel method to divide the instantaneous flow field into strictly monotonic elements is applied to tomographic particle-image velocimetry data.Especially the scaling regimes of velocity differences within each element with respect to their lengths are considered
Wall-shear stress measurements of turbulent flow over ribbed surfaces using the micro-pillar shear stress sensor MPS3
No full textThe drag reduction effect of a semi-circular riblet-structured surface in a turbulent boundary layer is experimentally investigated using the micro-pillar shear stress sensor MPS3. The MPS3 sensor is a novel tool for the quantitative measurement of the wall-shear stress distribution and possesses a high spatial and temporal resolution. The effectiveness and mechanisms of a ribbed surface in skin friction reduction are to be examined in comparison with the flow case of a flat surface
The University of Groningen in the world : a concise history /
No full textThe University of Groningen has been an international university since its foundation in 1614. The first professors formed a rich international community, and many students came from outside the Netherlands, especially from areas now belonging to Germany. Internationalization, a popular slogan nowadays, is therefore nothing new, but its meaning has changed over time. How did the University of Groningen grow from a provincial institution established for religious reasons into a top-100 university with 36,000 students, of whom 25% come from abroad and almost half of the academic staff is of foreign descent? What is the identity of this four-century-old university that is still strongly anchored in the northern part of the Netherlands but that also has a mind that is open to the world? The history of the university, as told by Klaas van Berkel and Guus Termeer, ends with a short paragraph on the impact of the corona crisis.Title from publisher's bibliographic system (viewed on 06 Dec 2022).The University of Groningen has been an international university since its foundation in 1614. The first professors formed a rich international community, and many students came from outside the Netherlands, especially from areas now belonging to Germany. Internationalization, a popular slogan nowadays, is therefore nothing new, but its meaning has changed over time. How did the University of Groningen grow from a provincial institution established for religious reasons into a top-100 university with 36,000 students, of whom 25% come from abroad and almost half of the academic staff is of foreign descent? What is the identity of this four-century-old university that is still strongly anchored in the northern part of the Netherlands but that also has a mind that is open to the world? The history of the university, as told by Klaas van Berkel and Guus Termeer, ends with a short paragraph on the impact of the corona crisis
Experimental investigation of turbulent boundary layer flow undergoing spanwise traveling transversal surface waves
No full textThe influence of spanwise traveling transversal surface waves on the near-wall flow field of turbulent boundary layers with and without adverse pressure gradient is investigated by particle-image velocimetry (PIV) and micro-particle tracking velocimetry (μ-PTV). A detailed analysis of the velocity profile immediately downstream of the actuated surface shows a local drag decrease as well as an increase caused by the surface wave dependent on the wave parameters
Uncertainty in perception and the Hierarchical Gaussian Filter.
Full text linkIn its full sense, perception rests on an agent's model of how its sensory input comes about and the inferences it draws based on this model. These inferences are necessarily uncertain. Here, we illustrate how the Hierarchical Gaussian Filter (HGF) offers a principled and generic way to deal with the several forms that uncertainty in perception takes. The HGF is a recent derivation of one-step update equations from Bayesian principles that rests on a hierarchical generative model of the environment and its (in)stability. It is computationally highly efficient, allows for online estimates of hidden states, and has found numerous applications to experimental data from human subjects. In this paper, we generalize previous descriptions of the HGF and its account of perceptual uncertainty. First, we explicitly formulate the extension of the HGF's hierarchy to any number of levels; second, we discuss how various forms of uncertainty are accommodated by the minimization of variational free energy as encoded in the update equations; third, we combine the HGF with decision models and demonstrate the inversion of this combination; finally, we report a simulation study that compared four optimization methods for inverting the HGF/decision model combination at different noise levels. These four methods (Nelder-Mead simplex algorithm, Gaussian process-based global optimization, variational Bayes and Markov chain Monte Carlo sampling) all performed well even under considerable noise, with variational Bayes offering the best combination of efficiency and informativeness of inference. Our results demonstrate that the HGF provides a principled, flexible, and efficient-but at the same time intuitive-framework for the resolution of perceptual uncertainty in behaving agents
- …
