On the performance of the flexible maximum entropy distributions within partially adaptive estimation

The partially adaptive estimation based on the assumed error distribution has emerged as a popular approach for estimating a regression model with non-normal errors. In this approach, if the assumed distribution is flexible enough to accommodate the shape of the true underlying error distribution, the efficiency of the partially adaptive estimator is expected to be close to the efficiency of the maximum likelihood estimator based on knowledge of the true error distribution. In this context, the maximum entropy distributions have attracted interest since such distributions have a very flexible functional form and nest most of the statistical distributions. Therefore, several flexible MaxEnt distributions under certain moment constraints are determined to use within the partially adaptive estimation procedure and their performances are evaluated relative to well-known estimators. The simulation results indicate that the determined partially adaptive estimators perform well for non-normal error distributions. In particular, some can be useful in dealing with small sample sizes. In addition, various linear regression applications with non-normal errors are provided.     

Designing shared spaces for firm‐community collaborations for innovation: Formal policies and coordination in open source projects

This paper adds to the growing debate on the design of the shared space between firms and communities in co‐innovation processes. We investigate specific organizational design initiatives that firms can take when collaborating with online communities to nurture their joint knowledge creation process and to realize a useful output. We contend that drafting a formal policy, leaving the coordination of the joint project to others and, in case of taking on the coordination duty, dedicating an employee's full time to realize the project administrator task can help increase the amount of useful output of such collaborations. Results based on a sample of 1,099 open source software projects hosted on the platform SourceForge.net support our hypotheses.     

Extracellular enzyme production and enterotoxigenic gene profiles of Bacillus cereus and Bacillus thuringiensis strains isolated from cheese in Turkey

The aim of the present study was to investigate the biochemical characteristics, extracellular enzyme production and enterotoxigenic genes contents of 6 Bacillus cereus and 22 Bacillus thuringiensis strains, isolated from 100 cheese samples in Turkey. Crystal morphologies of B. thuringiensis strains were found either spherical (n = 12) or spherical and irregular-shaped (n = 10) by phase contrast microscopy. B. cereus and B. thuringiensis strains were found to produce extracellular enzymes, respectively: gelatinase (83% and 91%), DNase (83% and 77%), lecithinase (83% and 95%), protease on skim milk agar (100% and 100%), protease on milk agar (100% and 91%), protease on casein agar (83% and 77%), xylanase (100% and 45%), and cellulase (0% and 41%), and amylase (83% and 27%). All of the strains, except for Bt-D1, hydrolyzed Tween 20 (96%), but not Tween 80 or tributyrin. Pectinolytic activity was obtained to be the least frequent (4%). PCR analysis showed that all strains contained nheA, nheB, nheC and hblD genes. The hblA and hblC genes were present in 2 and 4 of B. thuringiensis strains, respectively. The bceT gene was detected in 1 B. cereus and 9 B. thuringiensis strains. The entFM gene was detected more frequently in B. thuringiensis (82%) than in B. cereus strains (50%). To our knowledge, this is the first report about the isolation and identification of enterotoxigenic B. cereus and B. thuringiensis strains from cheese samples in Turkey.     

Performance Analysis of Turbo Product Codes in Multipath Visible Light Communication Systems

Wireless Personal Communications - Performance of visible light communication systems which is encoded using turbo product codes is studied in this article. It is the first time to show the effect...     

Gas chromatographic analyses of kerosene bioremediation displayed distinctive pattern of n-alkane degradation

Abstract

Kerosene, being one of the most commonly spilled petroleum products, needs to be removed from the polluted areas before it spreads over large habitats. Bioremediation is an environmental friendly approach for the cleanup of contaminated sites and exploits the potential of naturally occurring microbial populations or microorganisms with a known ability to degrade the contaminants. Therefore, it is important to get bacterial isolates with high kerosene degradation abilities for their potential use in kerosene bioremediation. For this reason, twenty previously identified hydrocarbon degraders were further analysed for their potential to degrade kerosene. Primary selection of kerosene degraders was done by using conventional enrichment culture containing 1% kerosene as a sole source of carbon. Gas chromatographic (GC) analyses of 17 pre-selected kerosene degraders displayed distinctive degradation pattern of n-alkanes including C₁₁-C₁₅, C₁₇ and C₁₉ fractions. The bacterial isolates Staphylococcus aureus Ba01, Delftia acidovorans Cd11, Acinetobacter calcoaceticus Fe10, Pseudomonas koreensis Hg11 and Acinetobacter johnsonii Sb01 sharing this distinctive degradation pattern displayed 70–84% degradation abilities in 21-day incubation. Among them, the isolates D. acidovorans Cd11 and A. johnsonii Sb01 with almost 81–84% degradation abilities and 0.08–0.09?mg/mL/day degradation rates showed the highest potential for kerosene bioremediation, respectively.     

Adsorptive polyHIPE composites based on biosorbent immobilized nanoclay: Effects of immobilization techniques

Styrene/divinyl benzene‐based macroporous polyHIPE composites were prepared from water‐in‐oil (w/o) high internal phase emulsion (HIPE) templates by using both organo‐modified montmorillonite (MMT) and a nonionic surfactant. For this purpose, Spirulina (Sp) microalgae was immobilized onto Na‐MMT clay by using two different modification techniques. They are based on conventional adsorption in solution (SOL) and novel cryoscopic expansion (C‐XP) assisted adsorption. Highly porous nanocomposites were prepared by using different percentages of modified nanoclays (SpSOLM/SpXPM) with a constant internal phase volume of 80%. The emulsion stability, morphology, and dye adsorption capacities were discussed by paying attention to nanoclay immobilization techniques, clay loading degree and surfactant concentration. The critical amount of nonionic surfactant for formation of the stable neat HIPE template was found to be only 5 vol% with respect to volume of organic phase. However, this amount was further reduced to much less value (2 vol%) with Sp immobilized nanoclays via help of cooperative interactions of Sp and MMT nanoclay. The C‐XP assisted modification of clay led to nanocomposites with 580% higher adsorption capacity for cationic dye. This remarkable benefit was obtained with even 0.5% clay loading and only 2% surfactant concentration. POLYM. ENG. SCI., 58:1229–1240, 2018. © 2017 Society of Plastics Engineers     

Fabrication of porous silica with controllable and tunable porosity via freeze casting

Porous ceramics offer unique properties that can bring advances to many application areas. The freeze-casting process has a strong potential for fabricating porous ceramics; however, the effects of process parameters on part porosity must be well understood for scalable manufacturing via freeze casting. This paper presents an experimental analysis of the freeze-casting process that correlates the freeze-casting parameters with pore characteristics. A full-factorial design of experiments is conducted on a unidirectional freeze-casting testbed using silica as the ceramic material and camphene as the solvent. The effects of solid loading, particle size, cooling temperature, and the distance from the cooling surface on porosity characteristics are evaluated. The fabricated samples are cross-sectioned vertically and horizontally and imaged using scanning electron microscopy. Image processing is used to obtain the porosity characteristics of areal porosity, pore size, pore shape, and pore orientation. The capability to steer the pore orientation is also demonstrated through bidirectional freezing experiments supported by a finite-element model. As a result, a quantitative understanding of the effects of freeze-casting process parameters on porosity characteristics is gained for the silica–camphene system. These results and the presented approach can be used for reproducible manufacture of porous ceramics with controlled porosity.     

Evaluation of Isolated Impact of Trans Fatty Acids on Short Dough Product

In this study, the effects of trans fatty acids (TFAs) on rheological properties of dough (elastic moduli (G’), loss moduli (G”), complex modulus (G*)) and textural properties of dough and cookie (hardness) were studied. Two different groups of fat samples having different TFA composition but similar solid fat content (SFC) were prepared. Samples of the first group (group 1) had TFA between 0.0 and 56.23 %, while the samples of the second group (group 2) contained trans isomers ranging from 0.0 to 44.40 %. Texture measurements of different doughs and cookies prepared with different fat samples were performed. Texture measurements indicated that hardness values of doughs increased from 3950 ± 420 to 5498 ± 506 g in group 1 and 4700 ± 501 to 6787 ± 369 g in group 2 with increased amounts of TFAs. A particularly high, almost three‐fold increase in complex modulus values was observed in the dough samples containing the highest TFA levels compared with samples containing 0.0 % TFA. Although not significantly different, mean hardness and relative sound intensity values of cookies displayed an initial decreasing trend and then both parameters had maximum values when the TFA content was at maximum in both groups.     

Reliability Testing of 3D-Printed Electromechanical Scanning Devices

Recent advances in the field of stereolithography based manufacturing, have led to a number of 3D-printed sensor and actuator devices, as a cost-effective and low fabrication complexity alternative to micro-electro-mechanical counterparts. Yet the reliability of such 3D-printed dynamic structures have yet to be explored. Here we perform reliability tests and analysis of a selected 3D-printed actuator, namely an electromechanical scanner. The scanner is targeted towards scanning incoming light onto the target, which is particularly useful for barcoding, display, and opto-medical tissue imaging applications. We monitor the deviations in the fundamental mechanical resonance, scan-line, and the quality factor on a number of scanners having different device thicknesses, for a total duration of 5 days (corresponding to 20–80 million cycles, depending on the device operating frequency). A total of 9 scanning devices, having 10 mm?×?10 mm die size were tested, with a highlight on device-device variability, as well as the effect of device thickness itself. An average standard deviation of < ~%10 (with respect to the mean) was observed for all tested parameters among scanners of the same type (an indicator device to device variability), while an average standard deviation of less than about 10 percent (with respect to the mean) was observed for all parameters for the duration of the entire test (as an indicator of device reliability), for a total optical scan angle of 5 degrees.     

Seizure onset detection based on frequency domain metric of empirical mode decomposition

This paper explores the data-driven properties of the empirical mode decomposition (EMD) for detection of epileptic seizures. A new method in frequency domain is presented to analyze intrinsic mode functions (IMFs) decomposed by EMD. They are used to determine whether the electroencephalogram (EEG) recordings contain seizure or not. Energy levels of the IMFs are extracted as threshold level to detect the changes caused by seizure activity. A scalar value energy resulting from the energy levels is individually used as an indicator of the epileptic EEG without the requirements of multidimensional feature vector and complex machine learning algorithms. The proposed methods are tested on different EEG recordings to evaluate the effectiveness of the proposed method and yield accuracy rate up to 97.89%.