Bubble load measurement in the pulp zone of industrial flotation machines - a new device for determining the froth recovery of attached particles

A new device has been developed to directly measure the bubble loading of particle-bubble aggregates in industrial flotation machines, both mechanical flotation cells as well as flotation column cells. The bubble loading of aggregates allows for in-depth analysis of the operating performance of a flotation machine in terms of both pulp/collection zone and froth zone performance. This paper presents the methodology along with an example showing the excellent reproducibility of the device and an analysis of different operating conditions of the device itself. (C) 2004 Elsevier B.V All rights reserved.

SVMtm: Support vector machines to predict transmembrane segments

A new method has been developed for prediction of transmembrane helices using support vector machines. Different coding schemes of protein sequences were explored, and their performances were assessed by crossvalidation tests. The best performance method can predict the transmembrane helices with sensitivity of 93.4% and precision of 92.0%. For each predicted transmembrane segment, a score is given to show the strength of transmembrane signal and the prediction reliability. In particular, this method can distinguish transmembrane proteins from soluble proteins with an accuracy of similar to99%. This method can be used to complement current transmembrane helix prediction methods and can be Used for consensus analysis of entire proteomes . The predictor is located at http://genet.imb.uq.edu.au/predictors/ SVMtm. (C) 2004 Wiley Periodicals, Inc.

Milling quality and protein properties of autumn-sown Chinese wheats evaluated through multi-location trials

Eight milling quality and protein properties of autumn-sown Chinese wheats were investigated using 59 cultivars and advanced lines grown in 14 locations in China from 1995 to 1998. Wide ranges of variability for all traits were observed across genotypes and locations. Genotype, location, year, and their interactions all significantly influenced most of the quality parameters. Kernel hardness, Zeleny sedimentation value, and mixograph development time were predominantly influenced by the effects of genotype. Genotype, location and genotype x location interaction were all important sources of variation for thousand kernel weight, test weight, protein content, and falling number, whereas genotype x location interaction had the largest effect on flour yield. Most of the genotypes were characterized by weak gluten strength with Zeleny sedimentation values less than 40 ml and mixograph development time shorter than 3 min. Eight groups of genotypes were recognized based on the average quality performance, grain hardness and gluten strength were the two parameters that determined the grouping, with contributions from protein content. Genotypes such as Zhongyou 16 and Annong 8903 displayed good milling quality, high grain hardness, protein content and strong gluten strength with high sedimentation value and long mixograph development time. Genotypes such as Lumai 15 and Yumai 18 were characterized by low grain hardness, protein content and weak gluten strength. Genotypes such as Yannong 15 and Chuanmai 24 were characterized by strong gluten strength with high sedimentation value and long mixograph development time, but low grain hardness and protein content lower than 12.3%. Genotypes such as Jingdong 6 and Xi'an 8 had weak gluten strength, but with high grain hardness and protein content higher than 12.2%. Five groups of locations were identified, and protein content and gluten strength were the two parameters that determined the grouping. Beijing, Shijiazhuang, Nanyang, Zhumadian and Nanjing produced wheats with medium to strong gluten strength and medium protein content, although there was still a large variation for most of the traits investigated between the locations. Wheat produced in Yantai was characterized by strong gluten strength, but with low protein content. Jinan, Anyang and Linfen locations produced wheats with medium to weak gluten strength and medium to high protein content. Wheats produced in Yangling, Zhenzhou, and Chengdu were characterized by weak gluten strength with medium to low protein content, whereas wheats produced in Xuzhou and Wuhan were characterized by weak gluten strength with low protein content. Industrial grain quality could be substantially improved through integrating knowledge of geographic genotype distribution with key location variables that affected end-use quality.

Hydrogen storage properties of MgH2/SWNT composite prepared by ball milling

A systematic investigation was performed on the hydrogen storage properties of mechano-chemically prepared MgH2/Single-walled carbon nanotube (SWNT) composites. It is found that the hydrogen absorption capacity and hydriding kinetics of the composites were dependent on the addition amount of SWNTs as well as milling time. A 5 wt.% addition of SVVNTs is optimum to facilitate the hydrogen absorption and desorption of MgH2. The composite MgH2/5 wt.% SWNTs milled for 10h can absorb 6.7 wt.% hydrogen within about 2 min at 573 K, and desorb 6 wt.% hydrogen in about 5 min at 623 K. Prolonging the milling time over 10 h leads to a serious degradation on hydrogen storage property of the MgH2/SWNT composite, and property/structure investigations suggest that the property degradation comes from the structure destruction of the SWNTs. (c) 2005 Elsevier B.V. All rights reserved.

Detecting and sorting targeting peptides wtih neural networks and support vector machines

This paper presents a composite multi-layer classifier system for predicting the subcellular localization of proteins based on their amino acid sequence. The work is an extension of our previous predictor PProwler v1.1 which is itself built upon the series of predictors SignalP and TargetP. In this study we outline experiments conducted to improve the classifier design. The major improvement came from using Support Vector machines as a "smart gate" sorting the outputs of several different targeting peptide detection networks. Our final model (PProwler v1.2) gives MCC values of 0.873 for non-plant and 0.849 for plant proteins. The model improves upon the accuracy of our previous subcellular localization predictor (PProwler v1.1) by 2% for plant data (which represents 7.5% improvement upon TargetP).

The effect of Cu addition and milling contaminations on the microstructure evolution of ball milled Al-Pb alloy during sintering

Al-10 wt.%Pb and Al-10 wt.%Pb-x wt.%Cu (x = 0-7.0) bulk alloys were prepared by sintering the mechanically alloyed powders at various temperatures. The microstructure changes of the as consolidated powders in the course of sintering were analyzed by differential scanning calorimetry, scanning electron microscopy, X-ray diffraction analysis and transmission electron microscopy. It has been found that, with respect to the Al-10 wt.%Pb-x wt.%Cu alloy, CuAl2 and Cu9Al4 phases formed in the milling process, and the amount of CuAl2 phase increased while the Cu9Al4 phase disappeared gradually in the sintering process. In both Al-10 wt.%Pb and Al-10 wt.%Pb-x wt.%Cu alloys, the sintering process results in the coarsening of Pb phase and the growth rate of Pb phase fulfills the Lifshitz-Slyozov-Wagner equation even though the size of the Pb phase was in nanometer range. The Pb particle exhibits cuboctahedral morphology and has a cubic to cubic orientation relationship with the Al matrix. The addition of Cu strongly depressed the growth rate of Pb. Contamination induced by milling has apparent influence on the microstructure of the sintered alloys. Al7Cu2Fe and aluminium oxide phases were identified in the sintered alloys. The cuboctahedral morphology of Pb particles was broken up by the presence of the oxide phase. (c) 2006 Elsevier B.V. All rights reserved.

Authorship Attribution with Support Vector Machines

In this paper we explore the use of text-mining methods for the identification of the author of a text. We apply the support vector machine (SVM) to this problem, as it is able to cope with half a million of inputs it requires no feature selection and can process the frequency vector of all words of a text. We performed a number of experiments with texts from a German newspaper. With nearly perfect reliability the SVM was able to reject other authors and detected the target author in 60–80% of the cases. In a second experiment, we ignored nouns, verbs and adjectives and replaced them by grammatical tags and bigrams. This resulted in slightly reduced performance. Author detection with SVMs on full word forms was remarkably robust even if the author wrote about different topics.