A dataset has been classified by some unknown classifier into two types of points. What were the most important factors in determining the classification outcome? In this work, we employ an axiomatic approach in order to uniquely characterize an infl
uence measure: a function that, given a set of classified points, outputs a value for each feature corresponding to its influence in determining the classification outcome. We show that our influence measure takes on an intuitive form when the unknown classifier is linear. Finally, we employ our influence measure in order to analyze the effects of user profiling on Googles online display advertising.
We consider the proof complexity of the minimal complete fragment, KS, of standard deep inference systems for propositional logic. To examine the size of proofs we employ atomic flows, diagrams that trace structural changes through a proof but ignore
logical information. As results we obtain a polynomial simulation
In this article, stability estimates are given for the determination of the zeroth-order bounded perturbations of the biharmonic operator when the boundary Neumann measurements are made on the whole boundary and on slightly more than half the boundar
y, respectively. For the case of measurements on the whole boundary, the stability estimates are of ln-type and for the case of measurements on slightly more than half of the boundary, we derive estimates that are of ln ln-type.
We investigate the break-up of Newtonian/viscoelastic droplets in a viscoelastic/Newtonian matrix under the hydrodynamic conditions of a confined shear flow. Our numerical approach is based on a combination of Lattice-Boltzmann models (LBM) and Finit
e Difference (FD) schemes. LBM are used to model two immiscible fluids with variable viscosity ratio (i.e. the ratio of the droplet to matrix viscosity); FD schemes are used to model viscoelasticity, and the kinetics of the polymers is introduced using constitutive equations for viscoelastic fluids with finitely extensible non-linear elastic dumbbells with Peterlins closure (FENE-P). We study both strongly and weakly confined cases to highlight the role of matrix and droplet viscoelasticity in changing the droplet dynamics after the startup of a shear flow. Simulations provide easy access to quantities such as droplet deformation and orientation and will be used to quantitatively predict the critical Capillary number at which the droplet breaks, the latter being strongly correlated to the formation of multiple neckings at break-up. This study complements our previous investigation on the role of droplet viscoelasticity (A. Gupta & M. Sbragaglia, {it Phys. Rev. E} {bf 90}, 023305 (2014)), and is here further extended to the case of matrix viscoelasticity.
Information flow analysis has largely ignored the setting where the analyst has neither control over nor a complete model of the analyzed system. We formalize such limited information flow analyses and study an instance of it: detecting the usage of
data by websites. We prove that these problems are ones of causal inference. Leveraging this connection, we push beyond traditional information flow analysis to provide a systematic methodology based on experimental science and statistical analysis. Our methodology allows us to systematize prior works in the area viewing them as instances of a general approach. Our systematic study leads to practical advice for improving work on detecting data usage, a previously unformalized area. We illustrate these concepts with a series of experiments collecting data on the use of information by websites, which we statistically analyze.
This paper is motivated by the fact that many systems need to be maintained continually while the underlying costs change over time. The challenge is to continually maintain near-optimal solutions to the underlying optimization problems, without crea
ting too much churn in the solution itself. We model this as a multistage combinatorial optimization problem where the input is a sequence of cost functions (one for each time step); while we can change the solution from step to step, we incur an additional cost for every such change. We study the multistage matroid maintenance problem, where we need to maintain a base of a matroid in each time step under the changing cost functions and acquisition costs for adding new elements. The online version of this problem generalizes online paging. E.g., given a graph, we need to maintain a spanning tree $T_t$ at each step: we pay $c_t(T_t)$ for the cost of the tree at time $t$, and also $| T_tsetminus T_{t-1} |$ for the number of edges changed at this step. Our main result is an $O(log m log r)$-approximation, where $m$ is the number of elements/edges and $r$ is the rank of the matroid. We also give an $O(log m)$ approximation for the offline version of the problem. These bounds hold when the acquisition costs are non-uniform, in which caseboth these results are the best possible unless P=NP. We also study the perfect matching version of the problem, where we must maintain a perfect matching at each step under changing cost functions and costs for adding new elements. Surprisingly, the hardness drastically increases: for any constant $epsilon>0$, there is no $O(n^{1-epsilon})$-approximation to the multistage matching maintenance problem, even in the offline case.
We obtain, by extensive direct numerical simulations, trajectories of heavy inertial particles in two-dimensional, statistically steady, homogeneous, and isotropic turbulent flows, with friction. We show that the probability distribution function $ma
thcal{P}(kappa)$, of the trajectory curvature $kappa$, is such that, as $kappa to infty$, $mathcal{P}(kappa) sim kappa^{-h_{rm r}}$, with $h_{rm r} = 2.07 pm 0.09$. The exponent $h_{rm r}$ is universal, insofar as it is independent of the Stokes number ($rm{St}$) and the energy-injection wave number. We show that this exponent lies within error bars of their counterparts for trajectories of Lagrangian tracers. We demonstrate that the complexity of heavy-particle trajectories can be characterized by the number $N_{rm I}(t,{rm St})$ of inflection points (up until time $t$) in the trajectory and $n_{rm I} ({rm St}) equiv lim_{ttoinfty} frac{N_{rm I} (t,{rm St})}{t} sim {rm St}^{-Delta}$, where the exponent $Delta = 0.33 pm0.02$ is also universal.
We study numerically the phase-ordering kinetics of the two-dimensional site-diluted Ising model. The data can be interpreted in a framework motivated by renormalization-group concepts. Apart from the usual fixed point of the non-diluted system, ther
e exist two disorder fixed points, characterized by logarithmic and power-law growth of the ordered domains. This structure gives rise to a rich scaling behavior, with an interesting crossover due to the competition between fixed points, and violation of superuniversality.
We report on the valence fluctuation of Ce in CeMo$_{2}$Si$_{2}$C as studied by means of magnetic susceptibility $chi(T)$, specific heat $C(T)$, electrical resistivity $rho(T)$ and x-ray absorption spectroscopy. Powder x-ray diffraction revealed that
CeMo$_{2}$Si$_{2}$C crystallizes in CeCr$_{2}$Si$_{2}$C-type layered tetragonal crystal structure (space group textit{P4/mmm}). The unit cell volume of CeMo$_{2}$Si$_{2}$C deviates from the expected lanthanide contraction, indicating non-trivalent state of Ce ions in this compound. The observed weak temperature dependence of the magnetic susceptibility and its low value indicate that Ce ions are in valence fluctuating state. The formal $L_{III}$ Ce valence in CeMo$_{2}$Si$_{2}$C $<$$widetilde{ u}$$>$ = 3.11 as determined from x-ray absorption spectroscopy measurement is well bellow the value $<$$widetilde{ u}$$> simeq$ 3.4 in tetravalent Ce compound CeO$_{2}$. The temperature dependence of specific heat does not show any anomaly down to 1.8 K which rules out any magnetic ordering in the system. The Sommerfeld coefficient obtained from the specific heat data is $gamma$ = 23.4 mJ/mol,K$^{2}$. The electrical resistivity follows the $T{^2}$ behavior in the low temperature range below 35 K confirming a Fermi liquid behavior. Accordingly both the Kadowaki Wood ratio $A/gamma^{2}$ and the Sommerfeld Wilson ratio $chi(0)/gamma$ are in the range expected for Fermi-liquid systems. In order to get some information on the electronic states, we calculated the band structure within the density functional theory, eventhough this approach is not able to treat 4f electrons accurately. The non-$f$ electron states crossing the Fermi level have mostly Mo 4d character. They provide the states with which the 4f sates are strongly hybridized, leading to the intermediate valent state.
We study the Higgs boson pair production through $e^+e^-$ collision in the noncommutative(NC) extension of the standard model using the Seiberg-Witten maps of this to the first order of the noncommutative parameter $Theta_{mu u}$. This process is fo
rbidden in the standard model with background space-time being commutative. We find that the cross section of the pair production of Higgs boson (of intermediate and heavy mass) at the future Linear Collider(LC) can be quite significant for the NC scale $Lambda$ lying in the range $0.5 - 1.0$ TeV. Finally, using the direct experimental(LEP II, Tevatron and global electro-weak fit) bound on Higgs mass, we obtain bounds on the NC scale as 665 GeV $le Lambda le 998$ GeV.
