Ensemble of Efficient Minimal Learning Machines for Classification and Regression

Minimal Learning Machine (MLM) is a recently proposed supervised learning algorithm with performance comparable to most state-of-the-art machine learning methods. In this work, we propose ensemble methods for classification and regression using MLMs. The goal of ensemble strategies is to produce more robust and accurate models when compared to a single classifier or regression model. Despite its successful application, MLM employs a computationally intensive optimization problem as part of its test procedure (out-of-sample data estimation). This becomes even more noticeable in the context of ensemble learning, where multiple models are used. Aiming to provide fast alternatives to the standard MLM, we also propose the Nearest Neighbor Minimal Learning Machine and the Cubic Equation Minimal Learning Machine to cope with classification and single-output regression problems, respectively. The experimental assessment conducted on real-world datasets reports that ensemble of fast MLMs perform comparably or superiorly to reference machine learning algorithms.     

Molecular epidemiology and distribution of antimicrobial resistance genes of Staphylococcus species isolated from Chinese dairy cows with clinical mastitis

Staphylococcus species, categorized into Staphylococcus aureus and non-aureus staphylococci (NAS), are frequent causes of mastitis in dairy cattle around the world. Current treatments using antimicrobials are under increasing scrutiny due to rising prevalence of multi-drug resistance in S. aureus. Objectives of this study were to determine: (1) genetic diversity of Staphylococcus species isolated from clinical mastitis in cows from large Chinese dairy farms; and (2) prevalence and distribution of antimicrobial resistance genes (ARG) in these isolates. Staphylococcus aureus (n = 96) were isolated from 26 herds located in 12 provinces of China, whereas NAS (n = 112) were isolated from 59 herds located in 18 provinces of China. The NAS were identified at the species level using a partial 16S rRNA sequencing method, whereas random amplification of polymorphic DNA (RAPD) PCR was done to determine genetic relationships of isolates. Finally, PCR was used to detect resistance and biofilm formation genes. Staphylococcus chromogenes (33%) was the most common NAS species, followed by Staphylococcus sciuri (17%) and Staphylococcus epidermidis (8%). Staphylococcus aureus was grouped in 12 genotypes, of which 2 types represented 56% of isolates. Staphylococcus chromogenes (n = 37) clustered into 8 RAPD types, with 2 prevalent types containing 73% of isolates. The most prevalent ARG in S. aureus isolates was blaZ (95%), followed by tetM (33%), tetK (31%), ermT (26%), and aacA-aphD (23%). The mecA and vanA were detected in 16 and 4% of isolates, respectively. In NAS, blaZ (100%), mecA (73%), tetK (79%), tetM (96%), mphC (63%), and msrA (54%) were frequently detected. Antimicrobial resistance genes mecA, tetK, tetL, tetM, dfrG, ermB, msrA, mphC, aadD, and aphA3 were more commonly detected in NAS than in S. aureus. Biofilm formation genes (icaA and icaD) were frequently detected in staphylococci isolated from bovine clinical mastitis. The existence of predominant RAPD types in S. aureus and S. chromogenes isolates across Chinese dairy farms indicated that specific genotypes had disseminated within herds and become more udder-adapted. High prevalence of ARG, especially in NAS, highlighted the risk of selection of multi-drug resistant staphylococci with potential as a reservoir of ARG.     

Genetic diversity and molecular epidemiology of outbreaks of Klebsiella pneumoniae mastitis on two large Chinese dairy farms

Klebsiella pneumoniae is an opportunistic and environmental mastitis-causing pathogen, with potential for contagious transmission. Repetitive element sequence-based PCR was used to determine genetic diversity and explore potential transmission and reservoirs for mastitis caused by K. pneumoniae on 2 large Chinese dairy farms. A total of 1,354 samples was collected from the 2 dairy farms, including milk samples from cows with subclinical and clinical mastitis, bedding, feces, feed, teat skin, and milking liners. Environmental samples were collected from all barns and milking parlors and extramammary samples from randomly selected dairy cows on both farms. In total, 272 and 93 K. pneumoniae isolates were obtained from Farms A and B, respectively (with ~8K and 2K lactating cows, respectively). Isolation rates from clinical mastitis (CM), subclinical mastitis (SCM), and environmental or extramammary samples were 34, 23 and 37%, respectively for Farm A and 42, 3, and 34% for Farm B. The K. pneumoniae isolated from CM milk and extramammary or environmental sources had high genetic diversity (index of diversity >90%) on the 2 farms and from SCM on Farm A. However, on Farm B, 9 SCM isolates were classified as 2 genotypes, resulting in a relatively low index of diversity (Simpson's index of diversity = 0.39; 95% CI = 0.08–0.70). Genotypes of K. pneumoniae causing mastitis were commonly detected in feces, bedding, and milking liners (Farm A), or from teat skin, sawdust bedding, and feed (Farm B). Based on its high level of genetic diversity, we inferred K. pneumoniae was an opportunistic and environmental pathogen causing outbreaks of CM on these 2 large Chinese dairy farms. Nevertheless, that only a few genotypes caused SCM implied some strains had increased udder adaptability and a contagious nature or a common extramammary source. Finally, control of intramammary infections caused by K. pneumoniae on large Chinese dairy farms must consider farm-level predictors, as the 2 outbreaks had distinct potential environmental sources of infection.     

Current advances concerning the most cited metal ions doped bioceramics and silicate-based bioactive glasses for bone tissue engineering

Studies related to biomaterials that stimulate the repair of living tissue have increased considerably, improving the quality of many people's lives that require surgery due to traumatic accidents, bone diseases, bone defects, and reconstructions. Among these biomaterials, bioceramics and bioactive glasses (BGs) have proved to be suitable for coating materials, cement, scaffolds, and nanoparticles, once they present good biocompatibility and degradability, able to generate osteoconduction on the surrounding tissue. However, the role of biomaterials in hard tissue engineering is not restricted to a structural replacement or for guiding tissue regeneration. Nowadays, it is expected that biomaterials develop a multifunctional role when implanted, orchestrating the process of tissue regeneration and providing to the body the capacity to heal itself. In this way, the incorporation of specific metal ions in bioceramics and BGs structure, including magnesium, silver, strontium, lithium, copper, iron, zinc, cobalt, and manganese are currently receiving enhanced interest as biomaterials for biomedical applications. When an ion is incorporated into the bioceramic structure, a new category of material is created, which has several unique properties that overcome the disadvantages of primitive material and favors its use in different biomedical applications. The doping can enhance handling properties, angiogenic and osteogenic performance, and antimicrobial activity. Therefore, this review aims to summarize the effect of selected metal ion dopants into bioceramics and silicate-based BGs in bone tissue engineering. Furthermore, new applications for doped bioceramics and BGs are highlighted, including cancer treatment and drug delivery.     

Thermal and mechanical properties of PES/PTFE composites and nanocomposites

Polyethersulphone/polytetrafluoroethylene (PES/PTFE) nanocomposites and composites were prepared by precipitation of PES into a PTFE latex‐containing nanoparticles. Different samples were obtained by varying the relative ratio between PES and PTFE. The complex crystallization process, discussed within the fractionated crystallization frame, allowed to identify and quantify different dispersion degree of the PTFE nanoparticles within the PES matrix. The different samples were thus divided into nanocomposite and composites. The effect of crystalline PTFE domains on the mobility of PES was investigated and discussed. The dynamic‐mechanical behavior was explained in terms of the particle aggregation state. The mechanical properties of the PES/PTFE composites were found to depend on both the dispersion and the concentration of the PTFE nanoparticles. In the glassy state the stiffness of the materials was found to increase with the dispersion degree, resulting higher for the nanocomposite with respect to composites. On the contrary, in the rubbery state the modulus was found proportional to the PTFE nanoparticles concentration, resulting higher in the composites with respect to the nanocomposite. © 2013 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 130: 3624–3633, 2013     

Long-Memory Errors in Time Series Regressions with a Unit Root

This paper is concerned with estimation and inference in univariate time series regression with a unit root when the error sequence exhibits long-range temporal dependence. We consider generating mechanisms for the unit root process which include models with or without a drift term and we study the limit behavior of least squares statistics in regression models without drift and trend, with drift but no time trend, and with drift and time trend. We derive the limit distribution and rate of convergence of the ordinary least squares (OLS) estimator of the unit root, the intercept and the time trend in the three regression models and for the two different data-generating processes. The limiting distributions for the OLS estimator differ from those obtained under the hypothesis of weakly dependent errors not only in terms of the limiting process involved but also in terms of functional form. Further, we characterize the asymptotic behavior of both the t statistics for testing the unit root hypothesis and the t statistic for the intercept and time trend coefficients. We find that t ratios either diverge to infinity or collapse to zero. The limiting behavior of Phillips's Z _α and Z _t semiparametric corrections is also analyzed and found to be similar to that of standard Dickey– Fuller tests. Our results indicate that misspecification of the temporal dependence features of the error sequence produces major effects on the asymptotic distribution of estimators and t ratios and suggest that alternative approaches might be more suited to testing for a unit root in time series regression.     

Temperature stimuli-responsive nanoparticles from chitosan-graft-poly(N-vinylcaprolactam) as a drug delivery system

This work describes the preparation of thermosensitive chitosan-graft-poly(N-vinylcaprolactam) nanoparticles by ionic gelation and their potential use as a controlled drug delivery system, using doxorubicin as a model drug. A systematic study of the effect of the main processing parameters on both the size and thermoresponsive behavior of nanoparticles was investigated. The size of the particles is strongly dependent on the length of the poly(N-vinylcaprolactam) grafted chains and the concentration of the copolymer and crosslinking agent solutions. The molecular structure of the copolymer plays an essential role in the phase transition temperature of the particles, which decreases with the length of PVCL grafted chain. The system displayed proper drug-association parameters, and the drug-loaded nanoparticles exhibited dose-dependent cytotoxicity. A significant increase in the doxorubicin delivery rate was observed above the phase transition temperature (40 °C). These features indicate that these nanoparticles are suitable for the development of a new thermally controlled anti-cancer drug delivery system. © 2019 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2019 , 136, 47831.