Adaptively preconditioned Krylov spaces to identify irrelevant predictors

Linear regression methods have problems in estimation when the predictor variables are highly correlated and when their number exceeds the number of available observations. PLS is one well known method for handling such ill-conditioned regression problems. It does so by approximating the regression solution in a low dimensional subspace. While it copes with collinearity and singularity problems, PLS does not have a variable selection procedure intrinsic to the method. However, it is often the case that one needs to decide which predictors, among the numerous and correlated ones, are the more relevant. The PLS coefficient is a good starting point for the identification of relevant variables in ill-conditioned regression settings. We propose to adaptively precondition the space generated by PLS in order to determine the most relevant predictors. The relevant subset is determined by a multiple testing procedure, and preconditioning stops when this set no longer changes. The principal objective is to do regression modelling and to recover solutions that are easy to interpret in the high dimensional regression setting. We use dimension reduction in a PLS fashion, using information on the response to guide the variable selection procedure. A variety of examples is studied with good results.     

Potato: a comparative study of the effect of cultivars and cultivation conditions and genetic modification on the physico-chemical properties of potato tubers in conjunction with multivariate analysis towards authenticity

Potato (Solanum tuberosum L.) is a highly nutritious, mild flavored, easy to blend food that has many possibilities for "building in" desired nutrients. Varietal and environmental differences are known to exist in the shape, size, and nutritional content of potatoes. Different populations opt for varying sensory properties in relation to their diets. Potatoes are a low energy food in comparison to cereals and legumes. The aim of this review was to present an update of the currently conducted studies both on the characterization of several potato varieties (physical, chemical, and sensory analysis) and by means of genetic modification. Towards this target, five comprehensive tables were compiled where all recent data (physicochemical properties) and GM varieties were presented in conjunction with multivariate analysis (chemometrics). The latter was shown to be effectively used towards authenticity purposes (identification of geographical origin, variety, GM).     

Optimizing bus services with variable directional and temporal demand using genetic algorithm

As a major mode choice of commuters for daily travel, bus transit plays an important role in many urban and metropolitan areas. This work proposes a mathematical model to optimize bus service by minimizing total cost and considering a temporally and directionally variable demand. An integrated bus service, consisting of all-stop and stop-skipping services is proposed and optimized subject to directional frequency conservation, capacity and operable fleet size constraints. Since the research problem is a combinatorial optimization problem, a genetic algorithm is developed to search for the optimal result in a large solution space. The model was successfully implemented on a bus transit route in the City of Chengdu, China, and the optimal solution was proved to be better than the original operation in terms of total cost. The sensitivity of model parameters to some key attributes/variables is analyzed and discussed to explore further the potential of accruing additional benefits or avoiding some of the drawbacks of stop-skipping services.     

A metaheuristic approach for the sequencing by hybridization problem with positive and negative errors

This work introduces a metaheuristic method for the reconstruction of the DNA string from its l-mer content in the presence of large amounts of positive and negative errors. The procedure consists of three parts: the formulation of the problem as an asymmetric traveling salesman problem (ATSP), a technique for handling the positive errors and an optimization algorithm that solves the formulated problem. The optimization algorithm is a variation of the threshold accepting method with intense local search and its function is controlled by a size diminishing shell. The optimization algorithm is used consecutively on ATSPs of continuously decreasing sizes till it reaches a final solution. The proposed method provides solutions of better quality compared to algorithms in the recent bibliography.     

Pneumatic artificial muscles: A switching Model Predictive Control approach

In this article, a switching Model Predictive Controller (sMPC) for a pneumatic artificial muscle (PAM) is presented. The control scheme is based on a switching PieceWise Affine (PWA) system model approximation that is able to capture the high nonlinearities of the PAM, while improving the overall model accuracy, and is composed of: (a) a feed-forward term regulating control input at specific reference set-points, and (b) a switching Model Predictive Controller handling any deviations from the system's equilibrium points. Extended experimental studies are being presented that prove the overall scheme's efficiency.     

Antioxidant and free radical-scavenging activities of phenolic extracts of olive fruits

The antioxidant ability of phenolic extracts of olive fruits during maturity in Chondrolia and Amfissis cultivars grown at 10 m and 300 m altitude showed that altitude affected phenol content, antioxidant and Fe²⁺/Fe³⁺ reduction/binding ability. The hydroxyl radical, peroxyl radical and peroxynitrite-induced DNA nicking assays have been used to evaluate the anti-radical activity of the extracts. Results showed that the ability of olive extracts to prevent radical-mediated DNA damage arises from the triple synergistic action of the genotype, the altitude and the maturation stage.     

Molecular and immunological characterization of Mus a 5 allergen from banana fruit

Scope Banana fruit has become an important cause of fruit allergy in the recent years. Among the five registered IUIS allergens, Mus a 1 and Mus a 2 have been characterized in detail. In this study, molecular characterization and evaluation of the allergenic properties of β‐1,3‐glucanase from banana (Musa acuminata), denoted as Mus a 5, were performed Methods and results The gene of Mus a 5 was cloned and sequenced. The obtained cDNA revealed a novel Mus a 5 isoform with an open reading frame encoding a protein of 340 amino acids comprising a putative signal peptide of 28 amino acid residues. By MALDI‐TOF analysis Mus a 5 isolated from banana fruit revealed a molecular mass of 33451±67 Da. Two Mus a 5 isoforms (pI 7.7 and 8.0) were detected by 2D immunoblot with an identical N‐terminal sequence. By mass fingerprint, 76 and 83% of the primary structure was confirmed for the two mature Mus a 5 isoforms, respectively. IgE reactivity to Mus a 5 was found in 74% of patients sensitized to banana fruit. Upregulation of basophil activation markers CD63 and CD203c was achieved with Mus a 5 in a concentration‐dependent manner Conclusion Mus a 5 is a functional allergen and a candidate for the component‐resolved allergy diagnosis of banana allergy.     

MULTI‐HOPPING INDUCED GAIN SCHEDULING FOR WIRELESS NETWORKED CONTROLLED SYSTEMS

A multi‐hopping induced gain scheduler for Wireless Networked Controlled Systems (WiNCS) is proposed. The control scheme is based on a client‐server architecture. The data packets are assumed to be carried over a network composed of wireless sensor nodes running the 802.11b protocol. The number of hops necessary for packets to reach their destinations is variable and depends on current traffic conditions. Changes in the number of hops induces changes in the latency times introduced by the communication network in the feedback loop. To deal with these changes an LQR‐output feedback scheme is introduced, whose parameters are adjusted according to the number of the hops. The weights of the LQR controllers are subsequently tuned using LMIs to ensure a prescribed stability margin despite the variable latency times. The overall scheme resembles a gain scheduled controller with the number of hops playing the role of the scheduling parameter. Simulation results on the NS‐2 simulator are provided to highlight the efficacy of the proposed scheme.