A new variational radial basis function approximation for inference in multivariate diffusions

In this paper we present a radial basis function based extension to a recently proposed variational algorithm for approximate inference for diffusion processes. Inference, for state and in particular (hyper-) parameters, in diffusion processes is a challenging and crucial task. We show that the new radial basis function approximation based algorithm converges to the original algorithm and has beneficial characteristics when estimating (hyper-)parameters. We validate our new approach on a nonlinear double well potential dynamical system.

Variational inference for diffusion processes

Diffusion processes are a family of continuous-time continuous-state stochastic processes that are in general only partially observed. The joint estimation of the forcing parameters and the system noise (volatility) in these dynamical systems is a crucial, but non-trivial task, especially when the system is nonlinear and multimodal. We propose a variational treatment of diffusion processes, which allows us to compute type II maximum likelihood estimates of the parameters by simple gradient techniques and which is computationally less demanding than most MCMC approaches. We also show how a cheap estimate of the posterior over the parameters can be constructed based on the variational free energy.

Market orientation of service networks:direct and indirect effects on sustained competitive advantage

Market orientation is an organization-wide concept that helps explain sustained competitive advantage (SCA). Since networks become ever more important, especially in the service sector, there is need to expand the concept of MO to a network setting. In line with Narver and Slater (1990), the concept of Market Orientation of Networks (MONW) is developed. This study indicates how MONW relates to the resource-based view (RBV) of the firm and the industrial organization (IO) view in explaining SCA. It is argued that MONW has direct and indirect effects on SCA. More precisely, the antecedent effect of MONW to resources and industry structure is considered.

Performance, emission and combustion characteristics of an indirect injection (IDI) multi-cylinder compression ignition (CI) engine operating on neat jatropha and karanj oils preheated by jacket water

Renewable non-edible plant oils such as jatropha and karanj have potential to substitute fossil diesel fuels in CI engines. A multi-cylinder water cooled IDI type CI engine has been tested with jatropha and karanj oils and comparisons made against fossil diesel. The physical and chemical properties of the three fuels were measured to investigate the suitability of jatropha and karanj oils as fuels for CI engines. The engine cooling water circuit and fuel supply systems were modified such that hot jacket water preheated the neat plant oil prior to injection. Between jatropha and karanj there was little difference in the performance, emission and combustion results. Compared to fossil diesel, the brake specific fuel consumption on volume basis was around 3% higher for the plant oils and the brake thermal efficiency was almost similar. Jatropha and karanj operation resulted in higher CO 2 and NO x emissions by 7% and 8% respectively, as compared to diesel. The cylinder gas pressure diagram showed stable engine operation with both plant oils. At full load, the plant oils gave around 3% higher peak cylinder pressure than fossil diesel. With the plant oils, cumulative heat release was smaller at low load and almost similar at full load, compared to diesel. At full load, the plant oils exhibited 5% shorter combustion duration. The study concludes that the IDI type CI engine can be efficiently operated with neat jatropha (or karanj) oil preheated by jacket water, after small modifications of the engine cooling and fuel supply circuits. © 2012 Elsevier Ltd.

Experimental investigation of performance, emission and combustion characteristics of an indirect injection multi-cylinder CI engine fuelled by blends of de-inking sludge pyrolysis oil with biodiesel

De-inking sludge can be converted into useful forms of energy to provide economic and environmental benefits. In this study, pyrolysis oil produced from de-inking sludge through an intermediate pyrolysis technique was blended with biodiesel derived from waste cooking oil, and tested in a multi-cylinder indirect injection type CI engine. The physical and chemical properties of pyrolysis oil and its blends (20 and 30 vol.%) were measured and compared with those of fossil diesel and pure biodiesel (B100). Full engine power was achieved with both blends, and very little difference in engine performance and emission results were observed between 20% and 30% blends. At full engine load, the brake specific fuel consumption on a volume basis was around 6% higher for the blends when compared to fossil diesel. The brake thermal efficiencies were about 3-6% lower than biodiesel and were similar to fossil diesel. Exhaust gas emissions of the blends contained 4% higher CO2 and 6-12% lower NOx, as compared to fossil diesel. At full load, CO emissions of the blends were decreased by 5-10 times. The cylinder gas pressure diagram showed stable engine operation with the 20% blend, but indicated minor knocking with 30% blend. Peak cylinder pressure of the 30% blend was about 5-6% higher compared to fossil diesel. At full load, the peak burn rate of combustion from the 30% blend was about 26% and 12% higher than fossil diesel and biodiesel respectively. In comparison to fossil diesel the combustion duration was decreased for both blends; for 30% blend at full load, the duration was almost 12% lower. The study concludes that up to 20% blend of de-inking sludge pyrolysis oil with biodiesel can be used in an indirect injection CI engine without adding any ignition additives or surfactants.

On an indirect integral equation approach for stationary heat transfer in semi-infinite layered domains in R3 with cavities

Regions containing internal boundaries such as composite materials arise in many applications.We consider a situation of a layered domain in IR3 containing a nite number of bounded cavities. The model is stationary heat transfer given by the Laplace equation with piecewise constant conductivity. The heat ux (a Neumann condition) is imposed on the bottom of the layered region and various boundary conditions are imposed on the cavities. The usual transmission (interface) conditions are satised at the interface layer, that is continuity of the solution and its normal derivative. To eciently calculate the stationary temperature eld in the semi-innite region, we employ a Green's matrix technique and reduce the problem to boundary integral equations (weakly singular) over the bounded surfaces of the cavities. For the numerical solution of these integral equations, we use Wienert's approach [20]. Assuming that each cavity is homeomorphic with the unit sphere, a fully discrete projection method with super-algebraic convergence order is proposed. A proof of an error estimate for the approximation is given as well. Numerical examples are presented that further highlights the eciency and accuracy of the proposed method.

Replication-based inference algorithms for hard computational problems

Inference algorithms based on evolving interactions between replicated solutions are introduced and analyzed on a prototypical NP-hard problem: the capacity of the binary Ising perceptron. The efficiency of the algorithm is examined numerically against that of the parallel tempering algorithm, showing improved performance in terms of the results obtained, computing requirements and simplicity of implementation. © 2013 American Physical Society.

Estimation and inference under economic restrictions

Estimation of economic relationships often requires imposition of constraints such as positivity or monotonicity on each observation. Methods to impose such constraints, however, vary depending upon the estimation technique employed. We describe a general methodology to impose (observation-specific) constraints for the class of linear regression estimators using a method known as constraint weighted bootstrapping. While this method has received attention in the nonparametric regression literature, we show how it can be applied for both parametric and nonparametric estimators. A benefit of this method is that imposing numerous constraints simultaneously can be performed seamlessly. We apply this method to Norwegian dairy farm data to estimate both unconstrained and constrained parametric and nonparametric models.

IPOs, clustering, indirect learning and filing independently

IPO underpricing has been attributed to valuation uncertainty, which can be at least partially resolved by the indirect learning associated with IPO clustering [Benveniste, L.M., Ljungqvist, A., Wilhelm, W.J., Yu, X.Y., 2003. Evidence of information spillovers in the production of investment banking services. Journal of Finance 58, 577–608]. We examine why firms might choose not to issue their IPOs contemporaneously with clusters of similar firms, forgoing opportunities to learn from their peers. We find that the willingness to file an IPO without the benefit of indirect learning from peer firm IPOs is directly related to insiders’ needs for portfolio diversification and the firm’s need to raise capital.

Message-passing for inference and optimization of real variables on sparse graphs

The inference and optimization in sparse graphs with real variables is studied using methods of statistical mechanics. Efficient distributed algorithms for the resource allocation problem are devised. Numerical simulations show excellent performance and full agreement with the theoretical results. © Springer-Verlag Berlin Heidelberg 2006.

A probabilistic indirect adaptive control for systems with input-dependent noise

A probabilistic indirect adaptive controller is proposed for the general nonlinear multivariate class of discrete time system. The proposed probabilistic framework incorporates input–dependent noise prediction parameters in the derivation of the optimal control law. Moreover, because noise can be nonstationary in practice, the proposed adaptive control algorithm provides an elegant method for estimating and tracking the noise. For illustration purposes, the developed method is applied to the affine class of nonlinear multivariate discrete time systems and the desired result is obtained: the optimal control law is determined by solving a cubic equation and the distribution of the tracking error is shown to be Gaussian with zero mean. The efficiency of the proposed scheme is demonstrated numerically through the simulation of an affine nonlinear system.

Variational mean-field algorithm for efficient inference in large systems of stochastic differential equations

This work introduces a Gaussian variational mean-field approximation for inference in dynamical systems which can be modeled by ordinary stochastic differential equations. This new approach allows one to express the variational free energy as a functional of the marginal moments of the approximating Gaussian process. A restriction of the moment equations to piecewise polynomial functions, over time, dramatically reduces the complexity of approximate inference for stochastic differential equation models and makes it comparable to that of discrete time hidden Markov models. The algorithm is demonstrated on state and parameter estimation for nonlinear problems with up to 1000 dimensional state vectors and compares the results empirically with various well-known inference methodologies.

Indirect ties in knowledge networks:a social network analysis with ordered weighted averaging operators

This PhD thesis analyses networks of knowledge flows, focusing on the role of indirect ties in the knowledge transfer, knowledge accumulation and knowledge creation process. It extends and improves existing methods for mapping networks of knowledge flows in two different applications and contributes to two stream of research. To support the underlying idea of this thesis, which is finding an alternative method to rank indirect network ties to shed a new light on the dynamics of knowledge transfer, we apply Ordered Weighted Averaging (OWA) to two different network contexts. Knowledge flows in patent citation networks and a company supply chain network are analysed using Social Network Analysis (SNA) and the OWA operator. The OWA is used here for the first time (i) to rank indirect citations in patent networks, providing new insight into their role in transferring knowledge among network nodes; and to analyse a long chain of patent generations along 13 years; (ii) to rank indirect relations in a company supply chain network, to shed light on the role of indirectly connected individuals involved in the knowledge transfer and creation processes and to contribute to the literature on knowledge management in a supply chain. In doing so, indirect ties are measured and their role as means of knowledge transfer is shown. Thus, this thesis represents a first attempt to bridge the OWA and SNA fields and to show that the two methods can be used together to enrich the understanding of the role of indirectly connected nodes in a network. More specifically, the OWA scores enrich our understanding of knowledge evolution over time within complex networks. Future research can show the usefulness of OWA operator in different complex networks, such as the on-line social networks that consists of thousand of nodes.