A new reverse reduce-error ensemble pruning algorithm

Although greedy algorithms possess high efficiency, they often receive suboptimal solutions of the ensemble pruning problem, since their exploration areas are limited in large extent. And another marked defect of almost all the currently existing ensemble pruning algorithms, including greedy ones, consists in: they simply abandon all of the classifiers which fail in the competition of ensemble selection, causing a considerable waste of useful resources and information. Inspired by these observations, an interesting greedy Reverse Reduce-Error (RRE) pruning algorithm incorporated with the operation of subtraction is proposed in this work. The RRE algorithm makes the best of the defeated candidate networks in a way that, the Worst Single Model (WSM) is chosen, and then, its votes are subtracted from the votes made by those selected components within the pruned ensemble. The reason is because, for most cases, the WSM might make mistakes in its estimation for the test samples. And, different from the classical RE, the near-optimal solution is produced based on the pruned error of all the available sequential subensembles. Besides, the backfitting step of RE algorithm is replaced with the selection step of a WSM in RRE. Moreover, the problem of ties might be solved more naturally with RRE. Finally, soft voting approach is employed in the testing to RRE algorithm. The performances of RE and RRE algorithms, and two baseline methods, i.e., the method which selects the Best Single Model (BSM) in the initial ensemble, and the method which retains all member networks of the initial ensemble (ALL), are evaluated on seven benchmark classification tasks under different initial ensemble setups. The results of the empirical investigation show the superiority of RRE over the other three ensemble pruning algorithms.     

A modified particle swarm optimization with multiple subpopulations for multimodal function optimization problems

In this paper, a modified particle swarm optimization (PSO) algorithm is developed for solving multimodal function optimization problems. The difference between the proposed method and the general PSO is to split up the original single population into several subpopulations according to the order of particles. The best particle within each subpopulation is recorded and then applied into the velocity updating formula to replace the original global best particle in the whole population. To update all particles in each subpopulation, the modified velocity formula is utilized. Based on the idea of multiple subpopulations, for the multimodal function optimization the several optima including the global and local solutions may probably be found by these best particles separately. To show the efficiency of the proposed method, two kinds of function optimizations are provided, including a single modal function optimization and a complex multimodal function optimization. Simulation results will demonstrate the convergence behavior of particles by the number of iterations, and the global and local system solutions are solved by these best particles of subpopulations.     

New nanocomposite based on poly(lactic-co-glycolic acid) copolymer and magnetite. Synthesis and characterization

The study presents the preparation of the new magnetic nanocomposite based on PLGA and magnetite. The PLGA used to obtain the magnetic nanocomposites was synthesized by the copolymerization of lactic acid with glycolic acid, in the presence of tin octanoate [Sn(Oct)₂] as catalyst, by polycondensation procedure. Magnetite was obtained by co-precipitation from aqueous salt solutions FeCl₂/FeCl₃. The particles size of magnetite was 420 nm, and the saturation magnetization 62.78 emu/g, while the PLGA/magnetite nanocomposite size was 864 nm and the saturation magnetization 39.44 emu/g. The magnetic nanocomposites were characterized by FT-IR, DLS technique, SEM, VSM and simultaneous thermal analyses (TG–FTIR–MS). The polymer matrix PLGA acts as a shell and carrier for the active component, while magnetite is the component which makes targeting possible by external magnetic field manipulation. Based on the gases resulted by thermal degradation of PLGA copolymer, using the simultaneous analysis TG–FTIR–MS, a possible degradation mechanism was proposed.     

Characteristics and formation mechanism of porosities in CFRP during laser joining of CFRP and steel

An experimental investigation on the mechanism of porosity formation during the laser joining of carbon fiber reinforced polymer (CFRP) and steel is presented. The porosity morphology and distribution were characterized by optical and scanning electron microscopy, and the thermal pyrolysis behaviors were investigated by thermal analysis and designed back-side cooling experiments. The results show that there are two types of porosities in CFRP. Porosity I only appears when the heat input is more than 77.8 J/mm. It has a smooth inner wall and distributes near the bonding interface between CFRP and steel at the central area of melted zone, which is caused by gaseous products such as CO₂, NH₃, H₂O, and hydrocarbons produced by the pyrolysis of CFRP. Porosity II can be seen under all joining conditions. It has a rough inner wall and distributes far away from the bonding interface, concentrating at the final solidification locations. Porosity II is caused by the shrinkage of melted CFRP during solidification stage.     

Combined hardware–software multi-parallel prefiltering on the Convey HC-1 for fast homology detection

Protein databases used in research are huge and still grow at a fast pace. Many comparisons need to be done when searching similar (homologous) sequences for a given query sequence in these databases. Comparing a query sequence against all sequences of a huge database using the well-known Smith–Waterman algorithm is very time-consuming. Hidden Markov Models pose an opportunity for reducing the number of entries of a database and also enable to find distantly homologous sequences. Fewer entries are achieved by clustering similar sequences in a Hidden Markov Model. Such an approach is used by the bioinformatics tool HHblits. To further reduce the runtime, HHblits uses two-level prefiltering to reduce the number of time-consuming Viterbi comparisons. Still, prefiltering is very time-consuming. Highly parallel architectures and huge bandwidth are required for processing and transferring the massive amounts of data. In this article, we present an approach exploiting the reconfigurable, hybrid computer architecture Convey HC-1 for migrating the most time-consuming part. The Convey HC-1 with four FPGAs and high memory bandwidth of up to 76.8 GB/s serves as the platform of choice. Other bioinformatics applications have already been successfully supported by the HC-1. Limited by FPGA size only, we present a design that calculates four first-level prefiltering scores per FPGA concurrently, i.e. 16 calculations in total. This score calculation for the query profile against database sequences is done by a modified Smith–Waterman scheme that is internally parallelized 128 times in contrast to the original Streaming ‘Single Instruction Multiple Data (SIMD)’ Extensions (SSE)-supported implementation where only 16-fold parallelism can be exploited and where memory bandwidth poses the limiting factor. Preloading the query profile, we are able to transform the memory-bound implementation to a compute- and resource-bound FPGA design. We tightly integrated the FPGA-based coprocessor into the hybrid computing system by employing task-parallelism for the two-level prefiltering. Despite much lower clock rates, the FPGAs outperform SSE-based execution for the calculation of the prefiltering scores by a factor of 7.9.     

Rapid detection of viable Bacillus cereus emetic and enterotoxic strains in food by coupling propidium monoazide and multiplex PCR (PMA-mPCR)

Bacillus cereus can cause emetic and diarrheal food poisoning. It is widespread in nature and therefore, considered a major foodborne pathogen. To develop a sensitive and reliable assay for detecting enterotoxin genes (nheA, entFM, hblD, cytK) and emetic toxin (ces), specific primers each targeting one individual gene were designed. Propidium monoazide (PMA) was coupled with the developed multiplex PCR (mPCR) for the detection of viable B. cereus. The inclusivity and exclusivity of the PMA-mPCR was confirmed using a panel of 44 strains including 17 emetic and 9 enterotoxic B. cereus reference strains and 18 non-target strains. The limit of detection (LOD) without PMA treatment in pure DNA was 2 pg/reaction tube. The LOD of mPCR assay in pure heat-killed dead bacteria was 4.0 × 10² CFU/mL. Also, the LOD on the viable bacteria with or without PMA treatment was similar (3.8 × 10² CFU/mL) showing that the PMA treatment did not significantly decrease sensitivity. Finally, the newly developed PMA-mPCR successfully detected 4.8 × 10³ and 3.6 × 10³ CFU/g of viable B. cereus F4810/72 (emetic) and B. cereus ATCC 12480 (enterotoxic) reference strains, respectively, in food samples. Hence, this study combines PMA and mPCR to detect viable B. cereus with a wide range of toxin detection (5 toxins). Thus, the novel PMA-mPCR assay developed in this study is a rapid and efficient diagnostic tool for the monitoring of viable B. cereus in food samples and potentially other samples via appropriate DNA extraction.     

Hydrocarbon-based electrode ionomer for proton exchange membrane fuel cells

The electrode ionomer is a key factor that significantly affects the catalyst layer morphology and fuel cell performance. Herein, sulfonated poly(arylene ether sulfone)-based electrode ionomers with polymers of various molecular weights and alcohol/water mixtures were prepared, and those comprising the alcohol/water mixture showed a higher performance than the ones prepared using higher boiling solvents, such as dimethylacetamide; this is owing to the formation of the uniformly dispersed ionomer catalyst layer. The relation between ionomer molecular weight for the same polymer structure and the sulfonation degree was investigated. Because the chain length of polymer varies with molecular weight and chain entanglement degree, its molecular weight affects the electrode morphology. As the ionomer covered the catalyst, the agglomerates formed were of different morphologies according to their molecular weight, which could be deduced indirectly through dynamic light scattering and scanning electron microscopy. Additionally, the fuel cell performance was confirmed in the current-voltage curve.     

Formulation of sage essential oil (Salvia officinalis,L.) monoterpenes into chitosan hydrogels and permeation study with GC-MS analysis

This study deals with the formulation of natural drugs into hydrogels. For the first time, compounds from the sage essential oil were formulated into chitosan hydrogels. A sample preparation procedure for hydrophobic volatile analytes present in a hydrophilic water matrix along with an analytical method based on the gas chromatography coupled with the mass spectrometry (GC-MS) was developed and applied for the evaluation of the identity and quantity of essential oil components in the hydrogels and saline samples. The experimental results revealed that the chitosan hydrogels are suitable for the formulation of sage essential oil. The monoterpene release can be effectively controlled by both chitosan and caffeine concentration in the hydrogels. Permeation experiment, based on a hydrogel with the optimized composition [3.5% (w/w) sage essential oil, 2.0% (w/w) caffeine, 2.5% (w/w) chitosan and 0.1% (w/w) Tween-80] in donor compartment, saline solution in acceptor compartment, and semi-permeable cellophane membrane, demonstrated the useful permeation selectivity. Here, (according to lipophilicity) an enhanced permeation of the bicyclic monoterpenes with antiflogistic and antiseptic properties (eucalyptol, camphor and borneol) and, at the same time, suppressed permeation of toxic thujone (not exceeding its permitted applicable concentration) was observed. These properties highlight the pharmaceutical importance of the developed chitosan hydrogel formulating sage essential oil in the dermal applications.     

Parallel chaos optimization algorithm with migration and merging operation

Chaos optimization algorithm (COA) utilizes the chaotic maps to generate the pseudo-random sequences mapped as the decision variables for global optimization applications. A kind of parallel chaos optimization algorithm (PCOA) has been proposed in our former studies to improve COA. The salient feature of PCOA lies in its pseudo-parallel mechanism. However, all individuals in the PCOA search independently without utilizing the fitness and diversity information of the population. In view of the limitation of PCOA, a novel PCOA with migration and merging operation (denoted as MMO-PCOA) is proposed in this paper. Specifically, parallel individuals are randomly selected to be conducted migration and merging operation with the so far parallel solutions. Both migration and merging operation exchange information within population and produce new candidate individuals, which are different from those generated by stochastic chaotic sequences. Consequently, a good balance between exploration and exploitation can be achieved in the MMO-PCOA. The impacts of different one-dimensional maps and parallel numbers on the MMO-PCOA are also discussed. Benchmark functions and parameter identification problems are used to test the performance of the MMO-PCOA. Simulation results, compared with other optimization algorithms, show the superiority of the proposed MMO-PCOA algorithm.     

Mechanical damage characterization of glass fiber-reinforced polymer laminates by ultrasonic maps

A method for simultaneous measurement of the thickness and density for Glass Fiber-Reinforced Polymer (GFRP) laminate plates with ultrasonic waves in C-Scan mode is presented in the form of maps. The method uses three different signals in immersion pulse-echo C-Scan mode. The maps obtained based on the density show the heterogeneity of the material at high resolution at the pixel level (1 × 1 mm²) and therefore they represent an efficient tool to assess and evaluate the damage of the composite structures after manufacturing and after an applied mechanical loading.