3499 research outputs found
Sort by
Binary and multi-class Parkinsonian disorders classification using Support Vector Machines with graph-based features
No full textAn extreme value analysis of the last century crises across industries in the U.S. economy
No full textThe two large scale crises that hit the world economy in the last century, i.e. the Great
Depression and the Great Recession, have similar outbreak and recovery patterns with
respect to several macroeconomic variables. In particular, the largest depressions are
likely to be accompanied by stock-market crashes. This study investigates the behavior
of the U.S. stock market before, during and after deep downturns, focusing particularly
on the tails of the return distribution. We develop two automatic procedures to identify
multiple change-points in the tail of financial time series as well as in the co-crash and
co-boom probabilities of different markets. We then apply our methodology to twelve
time series representative of the sectors of the U.S. economy. We find that regime
shifts in the lower tail of the distribution tend to co-occur before deep downturns. Our
results contribute to a better understanding of the origin and systemic nature of large
scale events to make policy interventions more timely and effective
Efficient Syntax-Driven Lumping of Differential Equations
No full textWe present an algorithm to compute exact aggregations of a class of systems of ordinary differential equations (ODEs). Our approach consists in an extension of Paige and Tarjan’s seminal solution to the coarsest refinement problem by encoding an ODE system into a suitable discrete-state representation. In particular, we consider a simple extension of the syntax of elementary chemical reaction networks because (i) it can express ODEs with derivatives given by polynomials of degree at most two, which are relevant in many applications in natural sciences and engineering; and (ii) we can build on two recently introduced bisimulations, which yield two complementary notions of ODE lumping. Our algorithm computes the largest bisimulations in O(r⋅s⋅logs)O(r⋅s⋅logs) time, where r is the number of monomials and s is the number of variables in the ODEs. Numerical experiments on real-world models from biochemistry, electrical engineering, and structural mechanics show that our prototype is able to handle ODEs with millions of variables and monomials, providing significant model reductions
Node-to-segment and node-to-surface interface finite elements for fracture mechanics
No full textThe topologies of existing interface elements used to discretize cohesive cracks are such that they can be used to compute the relative displacements (displacement discontinuities) of two opposing segments (in 2D) or of two opposing facets (in 3D) belonging to the opposite crack faces and enforce the cohesive traction–separation relation. In the present work we propose a novel type of interface element for fracture mechanics sharing some analogies with the node-to-segment (in 2D) and with the node-to-surface (in
3D) contact elements. The displacement gap of a node belonging to the finite element discretization of one crack face with respect to its projected point on the opposite face is used to determine the cohesive tractions, the residual vector and its consistent linearization for an implicit solution scheme. The following advantages with respect to classical interface finite elements are demonstrated: (i) non-matching finite element discretizations of the opposite crack faces is possible; (ii) easy modeling of cohesive cracks with non-propagating crack tips; (iii) the internal rotational equilibrium of the interface element is assured. Detailed examples are provided to show the usefulness of the proposed approach in nonlinear fracture mechanics problems
A Hybrid Model Predictive Control Approach to Attitude Control with Minimum-Impulse-Bit Thrusters
No full textThis paper studies an important aspect of attitude control for a launcher's upper stage: the minimum impulse bit (MIB), that is, the minimum torque that can be exerted by the thrusters. We model this effect using principles of hybrid systems theory and we design a hybrid model predictive control scheme for the attitude control of a launcher during its long coasting period, aiming at minimizing the number of thrusters' actuations. We apply the proposed methodology to a nonlinear model of a typical upper stage with multi-payload capability
Hierarchical organization of functional connectivity in the mouse brain: a complex network approach
No full textThis paper represents a contribution to the study of the brain functional connectivity from the perspective of complex networks theory. More specifically, we apply graph theoretical analyses to provide evidence of the modular structure of the mouse brain and to shed light on its hierarchical organization. We propose a novel percolation analysis and we apply our approach to the analysis of a resting-state functional MRI data set from 41 mice. This approach reveals a robust hierarchical structure of modules persistent across different subjects. Importantly, we test this approach against a statistical benchmark (or null model) which constrains only the distributions of empirical correlations. Our results unambiguously show that the hierarchical character of the mouse brain modular structure is not trivially encoded into this lower-order constraint. Finally, we investigate the modular structure of the mouse brain by computing the Minimal Spanning Forest, a technique that identifies subnetworks characterized by the strongest internal correlations. This approach represents a faster alternative to other community detection methods and provides a means to rank modules on the basis of the strength of their internal edges
Learning from successes and failures in pharmaceutical R&D
No full textIn this paper, we build a cumulative innovation model to understand the role of both success and failure in the learning dynamics that characterize pharmaceutical R&D. We test the prediction of our model by means of a unique dataset that combines patent information with R&D projects, thus distinguishing patents related to successfully marketed products from those covering candidate drugs that failed in clinical trials. Results confirm model predictions showing that patents associated with successfully completed projects receive more citations than those associated with failed projects. However, we also show that failed projects can be in turn cited more often than patents lacking clinical or preclinical information. We further explore the `black box' of innovation, providing evidence that both successes and failures contribute to R&D investment decisions and knowledge dynamics in science-driven sectors
A global/local approach for the prediction of the electric response of cracked solar cells in photovoltaic modules under the action of mechanical loads
No full textA numerical approach based on the finite element method to assess the impact of cracks in Silicon solar cells on the electric response of photovoltaic modules is proposed. A global coarse-scale finite element model of the composite laminate is used for carrying out the structural analysis. The computed displacements at the edges of each solar cell are passed via a projection scheme as boundary conditions to a 3D local fine-scale finite element model of the cells which accounts for cohesive cracks. The evaluated crack opening displacements along the crack faces are finally used as input to an electric model characterizing the grid line/solar cell ensemble. The identification of the relation between the localized electric resistance due to cracks and the crack opening, to be used as a constitutive model of cracks, is finally discussed in reference to experimental tests performed in the laboratory