Probabilistic Inference in Credal Networks: New Complexity Results

Credal networks are graph-based statistical models whose parameters take values in a set, instead of being sharply specified as in traditional statistical models (e.g., Bayesian networks). The computational complexity of inferences on such models depends on the irrelevance/independence concept adopted. In this paper, we study inferential complexity under the concepts of epistemic irrelevance and strong independence. We show that inferences under strong independence are NP-hard even in trees with binary variables except for a single ternary one. We prove that under epistemic irrelevance the polynomial-time complexity of inferences in credal trees is not likely to extend to more general models (e.g., singly connected topologies). These results clearly distinguish networks that admit efficient inferences and those where inferences are most likely hard, and settle several open questions regarding their computational complexity. We show that these results remain valid even if we disallow the use of zero probabilities. We also show that the computation of bounds on the probability of the future state in a hidden Markov model is the same whether we assume epistemic irrelevance or strong independence, and we prove an analogous result for inference in Naive Bayes structures. These inferential equivalences are important for practitioners, as hidden Markov models and Naive Bayes networks are used in real applications of imprecise probability.

New Complexity Results for MAP in Bayesian Networks

This paper presents new results for the (partial) maximum a posteriori (MAP) problem in Bayesian networks, which is the problem of querying the most probable state configuration of some of the network variables given evidence. It is demonstrated that the problem remains hard even in networks with very simple topology, such as binary polytrees and simple trees (including the Naive Bayes structure), which extends previous complexity results. Furthermore, a Fully Polynomial Time Approximation Scheme for MAP in networks with bounded treewidth and bounded number of states per variable is developed. Approximation schemes were thought to be impossible, but here it is shown otherwise under the assumptions just mentioned, which are adopted in most applications.

New Results for the MAP Problem in Bayesian Networks

This paper presents new results for the (partial) maximum a posteriori (MAP) problem in Bayesian networks, which is the problem of querying the most probable state configuration of some of the network variables given evidence. First, it is demonstrated that the problem remains hard even in networks with very simple topology, such as binary polytrees and simple trees (including the Naive Bayes structure). Such proofs extend previous complexity results for the problem. Inapproximability results are also derived in the case of trees if the number of states per variable is not bounded. Although the problem is shown to be hard and inapproximable even in very simple scenarios, a new exact algorithm is described that is empirically fast in networks of bounded treewidth and bounded number of states per variable. The same algorithm is used as basis of a Fully Polynomial Time Approximation Scheme for MAP under such assumptions. Approximation schemes were generally thought to be impossible for this problem, but we show otherwise for classes of networks that are important in practice. The algorithms are extensively tested using some well-known networks as well as random generated cases to show their effectiveness.

Thermodynamic Properties of Dichloromethane, Bromochloromethane, and Dibromomethane under Elevated Pressure: Experimental Results and SAFT-VR Mie Predictions

The speeds of sound in dibromomethane, bromochloromethane, and dichloromethane have been measured in the temperature range from 293.15 to 313.15 K and at pressures up to 100 MPa. Densities and isobaric heat capacities at atmospheric pressure have been also determined. Experimental results were used to calculate the densities and isobaric heat capacities as the function of temperature and pressure by means of a numerical integration technique. Moreover, experimental data at atmospheric pressure were then used to determine the SAFT-VR Mie molecular parameters for these liquids. The accuracy of the model has been then evaluated using a comparison of derived experimental high-pressure data with those predicted using SAFT. It was found that the model provide the possibility to predict also the isobaric heat capacity of all selected haloalkanes within an error up to 6%.

Nanoscale reinforced orthopaedic bone cement with graphene oxide - a possible solution to aseptic loosening of cemented hip implants? Preliminary results of an ongoing study

Hip replacement surgery is amongst the most common orthopaedic operations performed in the UK. Aseptic loosening is responsible for 40% of hip revision procedures. Aseptic loosening is a result of cement mantle fatigue. The aim of the current study is to analyse the effect of nanoscale Graphene Oxide (GO) on the mechanical properties of orthopaedic bone cement. Study Design A experimental thermal and mechanical analysis was conducted in a laboratory set up conforming to international standards for bone cement testing according to ISO 5583. Testing was performed on control cement samples of Colacryl bone cement, and additional samples reinforced with variable wt% of Graphene Oxide containing composites – 0.1%, 0.25%, 0.5% and 1.0% GO loading. Pilot Data Porosity demonstrated a linear relationship with increasing wt% loading compared to control (p<0.001). Thermal characterisation demonstrated maximal temperature during polymerization, and generated exotherm were inversely proportional to w%t loading (p<0.05) Fatigue strength performed on the control and 0.1 and 0.25%wt loadings of GO demonstrate increased average cycles to failure compared to control specimens. A right shift of the Weibull curve was demonstrated for both wt% available currently. Logistic regression analysis for failure demonstrated significant increases in number of cycles to failure for both specimens compared to a control (p<0.001). Forward Plan Early results convey positive benefits at low wt% loadings of GO containing bone cement. Study completion and further analysis is required in order to elude to the optimum w%t of GO which conveys the greatest mechanical advantage.

Vlasov-kinetic computer simulations of electrostatic waves in dusty plasmas: an overview of recent results

A series of numerical simulations based on a recurrence-free Vlasov kinetic model using kinetic phase point trajectories are presented. Electron-ion plasmas and three-component (electron-ion-dust) dusty or complex plasmas are considered, via independent simulations. Considering all plasma components modeled through a kinetic approach, the linear and nonlinear behavior of ion-acoustic excitations is investigated. Maxwellian and kappa-type (superthermal) distribution functions are assumed, as initial conditions, in separate simulations for the sake of comparison. The focus is on the parametric dependence of ion-acoustic waves on the electron-to-ion temperature ratio and on the dust concentration. © 2014 EDP Sciences, SIF, Springer-Verlag Berlin Heidelberg.

One-Year Safety of Olodaterol Once Daily via Respimat® in Patients with GOLD 2-4 Chronic Obstructive Pulmonary Disease: Results of a Pre-Specified Pooled Analysis

The novel long-acting β2-agonist olodaterol demonstrated an acceptable safety profile in short-term phase II clinical studies. This analysis of four randomized, double-blind, placebo-controlled, parallel-group, phase III studies (1222.11, NCT00782210; 1222.12, NCT00782509; 1222.13, NCT00793624; 1222.14, NCT00796653) evaluated the long-term safety of olodaterol once daily (QD) in a large cohort of patients with moderate to very severe (Global initiative for chronic Obstructive Lung Disease 2-4) chronic obstructive pulmonary disease (COPD). The studies compared olodaterol (5 or 10 μg) QD via Respimat®, formoterol 12 μg twice daily (BID) via Aerolizer® (1222.13 and 1222.14), and placebo for 48 weeks. Patients continued receiving background maintenance therapy, with ∼60% receiving concomitant cardiovascular therapy and 25% having a history of concomitant cardiac disease. Pre-specified analyses of pooled data assessed the adverse events (AEs) and serious AEs in the whole population, and in subgroups with cardiac disease, along with in-depth electrocardiogram and Holter monitoring. In total, 3104 patients were included in the safety analysis: 876 received olodaterol 5 μg, 883 received olodaterol 10 μg, 885 received placebos, and 460 received formoterol 12 μg BID. Overall incidence of on-treatment AEs (71.2%), serious AEs (16.1%), and deaths (1.7%) were balanced across treatment groups. Respiratory and cardiovascular AEs, including major adverse cardiac events, were reported at similar frequencies in placebo and active treatment groups. The safety profiles of both olodaterol 5 μg (marketed and registered dose) and 10 μg QD delivered via Respimat® are comparable to placebo and formoterol BID in this population, with no safety signals identified.

Absolute magnitudes and slope parameters of Pan-STARRS PS1 asteroids -preliminary results

We present the study of absolute magnitude (H) and slope parameter (G) of 170,000 asteroids observed by the Pan-STARRS1 telescope during the period of 15 months within its 3-year all-sky survey mission. The exquisite photometry with photometric errors below 0.04 mag and well-defined filter and photometric system allowed to derive H and G with statistical and systematic errors. Our new approach lies in the Monte Carlo technique simulating rotation periods, amplitudes, and colors, and deriving most-likely H, G and their systematic errors. Comparison of H_M by Muinonen's phase function (Muinonen et al., 2010) with the Minor Planet Center database revealed a negative offset of 0.22±0.29 meaning that Pan-STARRS1 asteroids are fainter. We showed that the absolute magnitude derived by Muinonen's function is systematically larger on average by 0.14±0.29 and by 0.30±0.16 when assuming fixed slope parameter (G=0.15, G_{12}=0.53) than Bowell's absolute magnitude (Bowell et al., 1989). We also derived slope parameters of asteroids of known spectral types and showed a good agreement with the previous studies within the derived uncertainties. However, our systematic errors on G and G_{12} are significantly larger than in previous work, which is caused by poor temporal and phase coverage of vast majority of the detected asteroids. This disadvantage will vanish when full survey data will be available and ongoing extended and enhanced mission will provide new data.