Resource
2015 EN
Michele Longo · Alessandra Mainini
We maximize the expected utility from terminal wealth for an HARA investorwhen the market price of risk is an unobservable random variable. We computethe optimal portfolio explicitly and explore the effects of learning bycomparing it with the corresponding myopic policy. In particular, we show that,for a market price of risk constant in sign, the ratio between the portfoliounder partial observation and its myopic counterpart increases with respect torisk tolerance. As a consequence, the absolute value of the partial observationcase is larger (smaller) than the myopic one if the investor is more (less)risk tolerant than the logarithmic investor. Moreover, our explicitcomputations enable to study in details the so called hedging demand induced bylearning about market price of risk.
Resource
2015 EN
Michele Stawowy
Artifact-centric models for business processes recently raised a lot ofattention, as they manage to combine structural (i.e. data related) withdynamical (i.e. process related) aspects in a seamless way. Many frameworksdeveloped under this approach, although, are not built explicitly for planning,one of the most prominent operations related to business processes. In thispaper, we try to overcome this by proposing a framework named Dynamic KnowledgeBases, aimed at describing rich business domains through DescriptionLogic-based ontologies, and where a set of actions allows the system to evolveby modifying such ontologies. This framework, by offering action rewriting andknowledge partialization, represents a viable and formal environment to developdecision making and planning techniques for DL-based artifact-centric businessdomains.
Resource
2015 EN
Luis Aparicio · Michele Cicoli · Bhaskar Dutta
+4 more
We study the phenomenology of the CMSSM/mSUGRA with non-thermal neutralinodark matter. Besides the standard parameters of the CMSSM we include thereheating temperature as an extra parameter. Imposing radiative electroweaksymmetry breaking with a Higgs mass around 125 GeV and no dark matteroverproduction, we contrast the scenario with different experimental boundsfrom colliders (LEP, LHC), cosmic microwave background (Planck), direct (LUX,XENON100, CDMS, IceCube) and indirect (Fermi) dark matter searches. The allowedparameter space is characterised by a Higgsino-like LSP with a mass around 300GeV. The observed dark matter abundance can be saturated for reheatingtemperatures around 2 GeV while larger temperatures require extranon-neutralino dark matter candidates and extend the allowed parameter space.Sfermion and gluino masses are in the few TeV region. These scenarios can beachieved in string models of sequestered supersymmetry breaking which avoidcosmological moduli problems and are compatible with gauge couplingunification. Astrophysics and particle physics experiments will fullyinvestigate this non-thermal scenario in the near future.
Resource
2015 EN
Michele Conforti · Samuel Fiorini · Kanstantsin Pashkovich
The cut dominant of a graph is the unbounded polyhedron whose points are allthose that dominate some convex combination of proper cuts. Minimizing anonnegative linear function over the cut dominant is equivalent to finding aminimum weight cut in the graph. We give a forbidden-minor characterization ofthe graphs whose cut dominant can be defined by inequalities with integercoefficients and right-hand side at most 2. Our result is related to theforbidden-minor characterization of TSP-perfect graphs by Fonlupt and Naddef(Math. Prog., 1992). We prove that our result implies theirs, with a shorterproof. Furthermore, we establish general properties of forbidden minors forright-hand sides larger than 2.
Resource
2015 EN
John Bulava · Michele Della Morte · Jochen Heitger
+1 more
We report on a non-perturbative computation of the renormalization factor Z_Aof the axial vector current in three-flavour O(a) improved lattice QCD withWilson quarks and tree-level Symanzik improved gauge action and also recall ourrecent determination of the improvement coefficient c_A. Our normalization andimprovement conditions are formulated at constant physics in a Schr\"odingerfunctional setup. The normalization condition exploits the full, massive axialWard identity to reduce finite quark mass effects in the evaluation of Z_A andcorrelators with boundary wave functions to suppress excited statecontributions in the pseudoscalar channel.
Resource
2015 EN
Pablo Ramos · Markos Papadakis · Michele Pavanello
We have carried out a thorough benchmark of the FDE-ET method for calculatinghole transfer couplings. We have considered 10 exchange-correlationfunctionals, 3 non-additive kinetic energy functionals and 3 basis sets.Overall, we conclude that with a 7% mean relative unsigned error, the PBEfunctional coupled with the PW91k non-additive Kinetic energy functional and aTZP basis set constitutes the most stable, and accurate level of theory forhole-transfer coupling calculations. The FDE-ET method is found to be anexcellent tool for computing diabatic couplings for hole transfer reactions.
Resource
2015 UN
Yongxin Chen · Tryphon Georgiou · Michele Pavon
We present an overview of our recent work on implementable solutions to theSchroedinger bridge problem and their potential application to optimaltransport and various generalizations.
Resource
2015 EN
Antonio Cammi · Matteo Zanetti · Davide Chiesa
+7 more
In this work, the characterization of the full-power steady state of theTRIGA Mark II nuclear reactor of the University of Pavia is performed bycoupling Monte Carlo (MC) simulation for neutronics with "Multiphysics" modelfor thermal-hydraulics. Neutronic analyses have been performed starting from aMC model of the entire reactor system, based on the MCNP5 code, that wasalready validated in fresh fuel and zero-power configuration (in which thermaleffects are negligible) using the available experimental data as benchmark. Inorder to describe the full-power reactor configuration, the temperaturedistribution in the core is necessary. To evaluate it, a thermal-hydraulicmodel has been developed, using the power distribution results from MCsimulation as input. The thermal-hydraulic model is focused on the core activeregion and takes into account sub-cooled boiling effects present at fullreactor power. The obtained temperature distribution is then introduced in theMC model and a benchmark analysis is carried out to validate the model in freshfuel and full-power configuration. The good agreement between experimental dataand simulation results concerning full-power reactor criticality, proves thereliability of the adopted methodology of analysis, both from neutronics andthermal-hydraulics perspective.
Resource
2015 EN
Richard A Schwartz · Gabriele Torre · Anna Maria Massone
+1 more
Saturation affects a significant rate of images recorded by the AtmosphericImaging Assembly on the Solar Dynamics Observatory. This paper describes acomputational method and a technological pipeline for the de-saturation of suchimages, based on several mathematical ingredients like ExpectationMaximization, image correlation and interpolation. An analysis of thecomputational properties and demands of the pipeline, together with anassessment of its reliability are performed against a set of data recorded fromthe Feburary 25 2014 flaring event.
Resource
2015 EN
Michele Benzi · Valeria Simoncini
We consider the numerical approximation of $f({\cal A})b$ where $b\in{\mathbbR}^{N}$ and $\cal A$ is the sum of Kronecker products, that is ${\cal A}=M_2\otimes I + I \otimes M_1\in{\mathbb R}^{N\times N}$. Here $f$ is a regularfunction such that $f({\cal A})$ is well defined. We derive a computationalstrategy that significantly lowers the memory requirements and computationalefforts of the standard approximations, with special emphasis on theexponential function, for which the new procedure becomes particularlyadvantageous. Our findings are illustrated by numerical experiments withtypical functions used in applications.