Linear response algorithms for approximate inference in graphical models

Belief propagation (BP) on cyclic graphs is an efficient algorithm for computing approximate marginal probability distributions over single nodes and neighboring nodes in the graph. However, it does not prescribe a way to compute joint distributions over pairs of distant nodes in the graph. In this article, we propose two new algorithms for approximating these pairwise probabilities, based on the linear response theorem. The first is a propagation algorithm that is shown to converge if BP converges to a stable fixed point. The second algorithm is based on matrix inversion. Applying these ideas to gaussian random fields, we derive a propagation algorithm for computing the inverse of a matrix.     

Partial state and unknown input estimation for time-delay systems

This article considers the problem of estimating a partial set of the state vector and/or unknown input vector of linear systems driven by unknown inputs and time-varying delay in the state variables. Three types of reduced-order observers, namely, observers with delays, observers without internal delays and delay-free observers are proposed in this article. Existence conditions and design procedures are presented for the determination of parameters for each case of observers. Numerical examples are presented to illustrate the design procedures.     

On Density-Based Data Streams Clustering Algorithms: A Survey简

Clustering data streams has drawn lots of attention in the last few years due to their ever-growing presence. Data streams put additional challenges on clustering such as limited time and memory and one pass clustering. Furthermore, discovering clusters with arbitrary shapes is very important in data stream applications. Data streams are infinite and evolving over time, and we do not have any knowledge about the number of clusters. In a data stream environment due to various factors, some noise appears occasionally. Density-based method is a remarkable class in clustering data streams, which has the ability to discover arbitrary shape clusters and to detect noise. Furthermore, it does not need the nmnber of clusters in advance. Due to data stream characteristics, the traditional density-based clustering is not applicable. Recently, a lot of density-based clustering algorithms are extended for data streams. The main idea in these algorithms is using density- based methods in the clustering process and at the same time overcoming the constraints, which are put out by data streanFs nature. The purpose of this paper is to shed light on some algorithms in the literature on density-based clustering over data streams. We not only summarize the main density-based clustering algorithms on data streams, discuss their uniqueness and limitations, but also explain how they address the challenges in clustering data streams. Moreover, we investigate the evaluation metrics used in validating cluster quality and measuring algorithms＇ performance. It is hoped that this survey will serve as a steppingstone for researchers studying data streams clustering, particularly density-based algorithms.     

Silver (Ag) as a novel masking material in glass etching for microfluidics applications

In this paper, we report the use of a single masking film for deep glass etching in hydrofluoric acid (HF). Thin film silver (Ag) is the key masking material in this work enabling a simple and low cost fabrication of microfluidic structures. The Ag film was deposited by evaporation and etched in a diluted nitric acid and de-ionized water solution at a ratio of 1:3. Surface morphology for different thicknesses of Ag film and its correlation to the maximum achievable etch depth is analyzed. AFM results shows low roughness values (<5 nm), indicating the Ag films are of smooth surface. With a 100 nm Ag film, a 220 μm etch depth in borosilicate glass substrates were produced and by further thickening the Ag to 300 nm, etch depths exceeding 300 μm were successfully achieved. SEM images show that thinner Ag films are of finer grains, potentially a source for pinholes formation where rapid penetration of HF along the grain boundaries peels off the Ag film from the glass surface. However, the Ag film was found not to react with HF. The process was demonstrated in the fabrication of cavities for integration with other microfluidic devices.     

Dietary magnesium oxide interactions with sodium bicarbonate on cows in early lactation

Thirty-six Holstein cows were fed a diet of 50% concentrate and 50% corn silage (dry matter) for 12 wk postpartum. Treatments were 0, .4, and .8% magnesium oxide with or without .8% sodium bicarbonate in a 2 X 3 factorial arrangement. Dry matter intake was not different among treatments although the combined buffers had higher intake. Milk production was higher for the .4% magnesium oxide treatments either alone or with sodium bicarbonate as compared with either the 0 or .8% magnesium oxide treatments. Ruminal pH was increased with addition of sodium bicarbonate either alone or together with magnesium oxide. Dietary addition of .4% magnesium oxide either alone or with sodium bicarbonate increased total volatile fatty acid in ruminal fluid. Propionate and valerate were depressed in both of the combined buffered diets. Fecal pH was increased with magnesium oxide addition either alone or with sodium bicarbonate. Increasing magnesium oxide increased magnesium in plasma. No interactions in animal performance were significant for the two buffers.     

Design of unknown input functional observers for nonlinear systems with application to fault diagnosis

This paper considers the design of low-order unknown input functional observers for robust fault detection and isolation of a class of nonlinear Lipschitz systems subject to unknown inputs. The proposed functional observers can be used to generate residual signals to detect and isolate actuator faults. By using the generalized inverse approach, the effect of the unknown inputs can be decoupled completely from the residual signals. Conditions for the existence and stability of reduced-order unknown input functional observer are derived. A design procedure for the generation of residual signals to detect and isolate actuator faults is presented using the proposed unknown-input observer-based approach. A numerical example is given to illustrate the proposed fault diagnosis scheme in nonlinear systems subject to unknown inputs.     

Solution-induced changes in the properties of polyethylenes

The chain entanglement states in high density, linear low density, and low density polyethylenes (HDPE, LLDPE, and LDPE) have been modified by recovering the polymers from solutions in trichlorobenzene (TCB) and p-xylene. In the thermodynamically good solvent, TCB, the entanglement density is assumed to be sharply reduced, a condition which is carried over to the corresponding solids. These display transient, but large increments in tensile moduli, slight changes in stress at rupture, and decreases in dynamic mechanical parameters and in elongation at rupture. Scanning calorimetry also shows these solids to have reduced crystallinity. Much smaller property modifications are noted in corresponding samples recovered from p-xylene. This liquid is a poorer solvent, particularly for HDPE and LLDPE. The results indicate that property modifications due to deliberate changes in the entanglement states of the polymers are a general phenomenon in the polyethylenes, and the magnitude of property changes depends on parameters of the molecular weight distribution. Chain branching does not seem to be a leading factor in the sensitivity of properties to modifications in the entanglement states. The property modifications produced by the present solution treatments are viewed as guides to the magnitude and duration of shear refining effects to be expected in HDPE, LLDPE, and LDPE polymers.     

A Variational Solution for Downdrag Force Analysis of Pile Groups

This article presents a variational approach for the analysis of downdrag forces of pile groups undergoing soil downward movement. The theoretical load‐transfer curves are employed to describe the soil load‐displacement relationship. A soil with stiffness varying with depth and with stiffer end‐bearing stratum at the pile toe can be easily treated using the present method. The validity of the present method is demonstrated through comparison with existing theoretical solutions and field measurements.     

Reliability analysis of laterally loaded piles using response surface methods

Response surface methods have been applied to the reliability analysis of laterally loaded piles. Beam elements and a series of discrete springs are used to model the pile–soil system. The pile head displacement and the maximum bending moment in the piles are used as the performance criteria in this study. It is shown in the illustrative example that the CDF and PDF curves of the pile head displacement and the maximum bending moment in the pile obtained from the proposed methods are in good agreement with Monte Carlo simulation. The failure probabilities of the pile under specified performance criteria, the probabilistic responses of the pile-soil system, and the effect of pile-soil parameters to the failure probability of the pile are also studied.     

Asphalt colloidal types differentiated by Korcak distribution

Asphalt in a colloidal context can be considered as a sol, sol-gel or gel type where in all three cases the asphaltene particle plays an important role. Usually the three colloidal types can be differentiated by refinery processes or by rheological properties. The aggregation or clustering of the asphaltene particles (or micelles) from sol to sol-gel and finally to the gel state with a fixed lattice can follow the stochastic process of fractal Brownian motion which can fit a hyperbolic (or Korcak) distribution. A test was carried out using the same mass of various asphalts of different colloidal types for floc formation with a series of n-alkanes (dispersion media) of increasing characteristic lengths. Results from floc formation studies indicate that the value of the size distribution for the gel type is small, is higher for the sol-gel type, and is highest for the sol type. On the other hand, the surface irregularities for all three types of asphalt do not have any correlation. These observed facts are consistent with principles based on the fractal approach.