Variance-covariance based weighing for neural network ensembles

Neural network (NN) is a popular artificial intelligence technique for solving complicated problems due to their inherent capabilities. However generalization in NN can be harmed by a number of factors including parameter's initialization, inappropriate network topology and setting parameters of the training process itself. Forecast combinations of NN models have the potential for improved generalization and lower training time. A weighted averaging based on Variance-Covariance method that assigns greater weight to the forecasts producing lower error, instead of equal weights is practiced in this paper. While implementing the method, combination of forecasts is done with all candidate models in one experiment and with the best selected models in another experiment. It is observed during the empirical analysis that forecasting accuracy is improved by combining the best individual NN models. Another finding of this study is that reducing the number of NN models increases the diversity and, hence, accuracy.

Validation of a food-frequency questionnaire assessment of carotenoid and vitamin E intake using weighed food records and plasma biomarkers: the method of triads model.

Background: Reliability or validity studies are important for the evaluation of measurement error in dietary assessment methods. An approach to validation known as the method of triads uses triangulation techniques to calculate the validity coefficient of a food-frequency questionnaire (FFQ).

Objective: To assess the validity of an FFQ estimates of carotenoid and vitamin E intake against serum biomarker measurements and weighed food records (WFRs), by applying the method of triads. Design: The study population was a sub-sample of adult participants in a randomised controlled trial of b-carotene and sunscreen in the prevention of skin cancer. Dietary intake was assessed by a self-administered FFQ and a WFR. Nonfasting blood samples were collected and plasma analysed for five carotenoids (a-carotene, b-carotene, b-cryptoxanthin, lutein, lycopene) and vitamin E. Correlation coefficients were calculated between each of the dietary methods and the validity coefficient was calculated using
the method of triads. The 95% confidence intervals for the validity coefficients were estimated using bootstrap sampling.

Results: The validity coefficients of the FFQ were highest for a-carotene (0.85) and lycopene (0.62), followed by b-carotene (0.55) and total carotenoids (0.55), while the lowest validity coefficient was for lutein (0.19). The method of triads could not be used for b-cryptoxanthin and vitamin E, as one of the three underlying correlations was negative.

Conclusions: Results were similar to other studies of validity using biomarkers and the method of triads. For many dietary factors, the upper limit of the validity coefficients was less than 0.5 and therefore only strong relationships between dietary exposure and disease will be detected.

Modelling for injection moulding process based on principal component regression method

Determination of the optimal operating condition for moulding process has been of special interest for many researchers. To determine the optimal setting, one has to derive the model of injection moulding process first which is able to map the relationship between the input process control factors and output responses. One of most popular modeling techniques is the linear least square regression due to its effectiveness and completeness. However, the least square regression was found to be very sensitive to the outliers and failed to provide a reliable model if the control variables are highly related with each other. To address this problem, a new modeling method based on principal component regression was proposed in this paper. The distinguished feature of our proposed method is it does not only consider the variance of covariance matrix of control variables but also consider the correlation coefficient between control variables and target variables to be optimised. Such a modelling method has been implemented into a commercial optimisation software and field test results demonstrated the performance of the proposed modelling method.

A modified moment-based image watermarking method robust to cropping attack

Developing a watermarking method that is robust to cropping attack is a challenging task in image watermarking. The moment-based watermarking schemes show good robustness to common signal processing attacks and some geometric attacks but are sensitive to cropping attack. In this paper, we modify the moment-based approach to deal with cropping attack. Firstly, we find the probability density function (pdf) of the pixel value distribution from the original image. Secondly, we reshape and normalize the pdf of the pixel value distribution (PPVD) to form a two dimensional image. Then, the moment invariants are calculated from the PPVD image. Since PPVD is insensitive to cropping, the proposed method is robust to cropping attack. Besides, it also has high robustness against other common attacks. Experimental results demonstrate the effectiveness of the proposed method.

A facile method to enhance ferroelectric properties in PVDF nanocomposites

Poly(vinylidene fluoride) (PVDF)/nanoclay composites were prepared using melt compounding. The effect of acrylic rubber (ACM) as a compatibilizer on different polymorph formation and on the ferroelectric properties of nanocomposites were investigated. The intercalation and morphological structure of the samples were studied using X-ray diffraction (XRD) and transmission electron microscopy (TEM). The infrared spectroscopy and X-ray analysis revealed the coexistence of β and γ crystalline forms in PVDF-clay nanocomposite, while in partially miscible PVDF/ACM/clay hybrids, three polymorphs of α, β and γ coexisted. The coefficients of electric field-polarization (E-P) Taylor expansion were calculated based on the Lorentz theory. Using a genetic algorithm, complex dielectric susceptibilities as well as the dielectric constants for each sample were calculated and optimized. The predicted dielectric constants were found to be in good agreement with the experimental results. A dielectric constant of 16 (10 Hz) was obtained for PVDF/ACM/clay (90/10/5), which was 40% higher than that of the PVDF-clay (100/5) nanocomposite without ACM. The improved dielectric performance of the nanocomposites can be attributed to the compatibilizing effect of ACM, which facilitated the growth of β polymorph in the sample.

An energy-aware QoS enhanced method for service computing across clouds and data centers

QoS plays a key role in evaluating a service or a service composition plan across clouds and data centers. Currently, the energy cost of a service's execution is not covered by the QoS framework, and a service's price is often fixed during its execution. However, energy consumption has a great contribution in determining the price of a cloud service. As a result, it is not reasonable if the price of a cloud service is calculated with a fixed energy consumption value, if part of a service's energy consumption could be saved during its execution. Taking advantage of the dynamic energy-Aware optimal technique, a QoS enhanced method for service computing is proposed, in this paper, through virtual machine (VM) scheduling. Technically, two typical QoS metrics, i.e., the price and the execution time are taken into consideration in our method. Moreover, our method consists of two dynamic optimal phases. The first optimal phase aims at dynamically benefiting a user with discount price by transparently migrating his or her task execution from a VM located at a server with high energy consumption to a low one. The second optimal phase aims at shortening task's execution time, through transparently migrating a task execution from a VM to another one located at a server with higher performance. Experimental evaluation upon large scale service computing across clouds demonstrates the validity of our method.

Fluid loss as a quick method to evaluate the hydraulic conductivity of geosynthetic clay liners under acidic conditions

This study investigates the performance of bentonite components of geosynthetic clay liners (GCLs) when exposed to aggressive leachates using the fluid loss test and provides a possible quick method for estimating the effect of acidic conditions on hydraulic conductivity. Fluid loss generally increases with increasing acid concentrations. Hydraulic conductivity values back-calculated from the fluid loss tests (kFL) are compared with the values measured using a flexible-wall permeameter (kTri).Generally, the predicted hydraulic conductivity values are conservative (kFL/kTri > 1) under water and low acid concentrations(≤0.015 mol/L). However, the back-calculated hydraulic conductivity is shown to be nonconservative (kFL/kTri < 1) at high acid concentrations (≥0.125 mol/L).

Modified moment-based image watermarking method robust to cropping attack

Developing a watermarking method that is robust to cropping attack is a challenging task in image watermarking. The moment-based watermarking schemes show good robustness to common signal processing attacks and some geometric attacks but are sensitive to cropping attack. In this paper, we modify the moment-based approach to deal with cropping attack. Firstly, we find the probability density function (PDF) of the pixel value distribution from the original image. Secondly, we reshape and normalize the pdf of the pixel value distribution (PPVD) to form a two dimensional image. Then, the moment invariants are calculated from the PPVD image. Since PPVD is insensitive to cropping, the proposed method is robust to cropping attack. Besides, it also has high robustness against other common attacks. Theoretical analysis and experimental results demonstrate the effectiveness of the proposed method.

Towards simulation of flapping wings using immersed boundary method

In this work the immersed boundary method is applied to simulate incompressible turbulent flows around stationary and moving objects. The goal is to demonstrate that the immersed boundary technique along with a large eddy simulation approach is capable of simulating the effect of the so-called leading edge vortex (LEV), which can be found in flapping wing aerodynamics. A Lagrangian method is used to approximatethe solutions in the freshly cleared cells that lay within solid objects at one time step and emerge into fluid domain at the next time step. Flow around a stationary cylinder at ReD D 20, 40, and 3900 (based oncylinder diameter D) is first studied to validate the immersed boundary solver based on the finite volume scheme using a staggered grid. Then, a harmonically oscillating cylinder at ReD D 10 000 is considered to test the solver after the Lagrangian method is implemented to interpolate the solution in the freshly cleared cells. Finally, this approach is used to study flows around a stationary flat-plate at several angles of attack and fast pitching flat-plate. The rapidly pitching plate creates a dynamic LEV that can be used to improve the efficiency of flapping wings of micro air vehicle and to determine the optimum flapping frequency.