A new RBF neural network with boundary value constraints

We present a novel topology of the radial basis function (RBF) neural network, referred to as the boundary value constraints (BVC)-RBF, which is able to automatically satisfy a set of BVC. Unlike most existing neural networks whereby the model is identified via learning from observational data only, the proposed BVC-RBF offers a generic framework by taking into account both the deterministic prior knowledge and the stochastic data in an intelligent manner. Like a conventional RBF, the proposed BVC-RBF has a linear-in-the-parameter structure, such that it is advantageous that many of the existing algorithms for linear-in-the-parameters models are directly applicable. The BVC satisfaction properties of the proposed BVC-RBF are discussed. Finally, numerical examples based on the combined D-optimality-based orthogonal least squares algorithm are utilized to illustrate the performance of the proposed BVC-RBF for completeness.

The effect of feeding modified soyabean oil enriched with C18 : 4 n-3 to broilers on the deposition of n-3 fatty acids in chicken meat

Supplementing broiler diets with conventional vegetable oils has little effect on the long-chain n-3 PUFA (LC n-3 PUFA) content of the meat. The present study investigated the effect on fatty acid composition and sensory characteristics of chicken meat when broilers were fed oil extracted from soyabeans (SDASOY) that had been genetically engineered to produce C18 : 4n-3 (stearidonic acid (SDA), 240 mg/g oil). Three diets were fed to 120 birds (eight replicate pens of five birds) from 15 d to slaughter (41–50 d). Diets were identical apart from the oil added to them (45 and 50 g/kg as fed in the grower and finisher phases, respectively), which was either SDASOY, near-isogenic soya (CON) or fish oil (FISH). The LC n-3 PUFA content of the meat increased in the order CON, SDASOY and FISH. In breast meat with skin, the SDA concentration was 522, 13 and 37 (sem 14·4) mg/100 g meat for SDASOY, CON and FISH, respectively. Equivalent values for C20 : 5n-3 (EPA) were 53, 13 and 140 (sem 8·4); for C22 : 5n-3 (docosapentaenoic acid (DPA)) 65, 15 and 101 (sem 3·5); for C22 : 6n-3 (DHA) 19, 9 and 181 (sem 4·4). Leg meat (with skin) values for SDA were 861, 23 and 68 (sem 30·1); for EPA 87, 9 and 258 (sem 7·5); for DPA 95, 20 and 165 (sem 5·0); for DHA 29, 10 and 278 (sem 8·4). Aroma, taste and aftertaste of freshly cooked breast meat were not affected. Fishy aromas, tastes and aftertastes were associated with LC n-3 PUFA content of the meat, being most noticeable in the FISH leg meat (both freshly cooked and reheated) and in the reheated SDASOY leg meat.

Givens rotation based fast backward elimination algorithm for RBF neural network pruning

A fast backward elimination algorithm is introduced based on a QR decomposition and Givens transformations to prune radial-basis-function networks. Nodes are sequentially removed using an increment of error variance criterion. The procedure is terminated by using a prediction risk criterion so as to obtain a model structure with good generalisation properties. The algorithm can be used to postprocess radial basis centres selected using a k-means routine and, in this mode, it provides a hybrid supervised centre selection approach.

Adaptive modelling with tunable RBF network using multi-innovation RLS algorithm assisted by swarm intelligence

In this paper, we propose a new on-line learning algorithm for the non-linear system identification: the swarm intelligence aided multi-innovation recursive least squares (SI-MRLS) algorithm. The SI-MRLS algorithm applies the particle swarm optimization (PSO) to construct a flexible radial basis function (RBF) model so that both the model structure and output weights can be adapted. By replacing an insignificant RBF node with a new one based on the increment of error variance criterion at every iteration, the model remains at a limited size. The multi-innovation RLS algorithm is used to update the RBF output weights which are known to have better accuracy than the classic RLS. The proposed method can produces a parsimonious model with good performance. Simulation result are also shown to verify the SI-MRLS algorithm.

A combined SMOTE and PSO based RBF classifier for two-class imbalanced problems

This contribution proposes a powerful technique for two-class imbalanced classification problems by combining the synthetic minority over-sampling technique (SMOTE) and the particle swarm optimisation (PSO) aided radial basis function (RBF) classifier. In order to enhance the significance of the small and specific region belonging to the positive class in the decision region, the SMOTE is applied to generate synthetic instances for the positive class to balance the training data set. Based on the over-sampled training data, the RBF classifier is constructed by applying the orthogonal forward selection procedure, in which the classifier's structure and the parameters of RBF kernels are determined using a PSO algorithm based on the criterion of minimising the leave-one-out misclassification rate. The experimental results obtained on a simulated imbalanced data set and three real imbalanced data sets are presented to demonstrate the effectiveness of our proposed algorithm.

Effect of dietary supplementation with selenium-enriched yeast or sodium selenite on selenium tissue distribution and meat quality in commercial-line turkeys

The objective of this study was to determine the concentration of total selenium (Se) and proportions of total Se comprised as selenomethionine (SeMet) and selenocysteine (SeCys) in the tissues of female turkeys offered diets containing graded additions of selenized-enriched yeast (SY), or sodium selenite (SS). Oxidative stability and tissue glutathione peroxidase (GSH-Px) activity of breast and thigh muscle were assessed at 0 and 10 days post mortem. A total of 216 female turkey poults were enrolled in the study. A total of 24 birds were euthanized at the start of the study and samples of blood, breast, thigh, heart, liver, kidney and gizzard were collected for determination of total Se. Remaining birds were blocked by live weight and randomly allocated to one of four dietary treatments(n548 birds/treatment) that differed either in Se source (SY v. SS) or dose (Con [0.2 mg/kg total Se], SY-L and SS-L [0.3mg/kg total Se as SY and SS, respectively] and SY-H [0.45mg total Se/kg]). Following 42 and 84 days of treatment 24 birds per treatment were euthanized and samples of blood, breast, thigh, heart, liver, kidney and gizzard were retained for determination of total Se and the proportion of total Se comprised as SeMet or SeCys. Whole blood GSH-Px activity was determined at each time point. Tissue GSH-Px activity and thiobarbituric acid reactive substances were determined in breast and thigh tissue at the end of the study. There were responses (P,0.001) in all tissues to the graded addition of dietary Se, although rates of accumulation were highest in birds offered SY. There were notable differences between tissue types and treatments in the distribution of SeMet and SeCys, and the activity of tissue and erythrocyte GSH-Px (P,0.05). SeCys was the predominant form of Se in visceral tissue and SeMet the predominant form in breast tissue. SeCys contents were greater in thigh when compared with breast tissue. Muscle tissue GSH-Px activities mirrored SeCys contents. Despite treatment differences in tissue GSH-Px activity, there were no effects of treatment on any meat quality parameter.

Blood pressure lowering effects of beetroot juice and novel beetroot enriched bread products in normotensive male subjects

A number of vegetables have a high nitrate content which after ingestion can be reduced to 36 nitrite by oral bacteria, and further to vasoprotective nitric oxide endogenously. Two separate 37 randomly controlled, single blind, cross-over, postprandial studies were performed in 38 normotensive volunteers. Ambulatory blood pressure was measured over a 24 h period 39 following consumption of either four doses of beetroot juice (BJ) 0 g, 100 g, 250 g and 500 g 40 (n = 18) or three bread products, control bread (0 g beetroot), red beetroot and white beetroot 41 enriched breads (n =14). Total urinary nitrate/nitrite (NOx) was measured at baseline, 2, 4 42 and 24 h post ingestion. BJ consumption significantly, and in a near dose dependent manner, 43 lowered systolic (P <0.01) and diastolic BP (P <0.001) over a period of 24 h, compared to 44 water control. Furthermore, bread products enriched with 100 g red or white beetroot lowered 45 systolic and diastolic BP over a period of 24 h (red beetroot enriched bread, P <0.05), with no 46 statistical differences between varieties. Total urinary NOx significantly increased following 47 consumption of 100 g (P<0.01), 250 g (P <0.001) and 500 g BJ (P <0.001) and after red 48 beetroot bread (P <0.05), but did not reach significance for white beetroot bread compared to 49 the no beetroot condition. These studies demonstrated significant hypotensive effects of a low 50 dose (100 g) of beetroot which was unaffected by processing, or the presence of betacyanins. 51 This data strengthens the evidence for cardioprotective BP lowering effects of dietary nitrate-52 rich vegetables.

Stable Nonlinear Receding Horizon Regulator Using RBF Neural Network Models

The general stability theory of nonlinear receding horizon controllers has attracted much attention over the last fifteen years, and many algorithms have been proposed to ensure closed-loop stability. On the other hand many reports exist regarding the use of artificial neural network models in nonlinear receding horizon control. However, little attention has been given to the stability issue of these specific controllers. This paper addresses this problem and proposes to cast the nonlinear receding horizon control based on neural network models within the framework of an existing stabilising algorithm.

A randomised, double-blind, placebo controlled cross-over study to determine the gastrointestinal effects of consumption of arabinoxylanoligosaccharides enriched bread in healthy volunteers

BACKGROUND: Prebiotics are food ingredients, usually non-digestible oligosaccharides, that are selectively fermented by populations of beneficial gut bacteria. Endoxylanases, altering the naturally present cereal arabinoxylans, are commonly used in the bread industry to improve dough and bread characteristics. Recently, an in situ method has been developed to produce arabinoxylan-oligosaccharides (AXOS) at high levels in breads through the use of a thermophilic endoxylanase. AXOS have demonstrated potentially prebiotic properties in that they have been observed to lead to beneficial shifts in the microbiota in vitro and in murine, poultry and human studies. METHODS: A double-blind, placebo controlled human intervention study was undertaken with 40 healthy adult volunteers to assess the impact of consumption of breads with in situ produced AXOS (containing 2.2 g AXOS) compared to non-endoxylanase treated breads. Volatile fatty acid concentrations in faeces were assessed and fluorescence in situ hybridisation was used to assess changes in gut microbial groups. Secretory immunoglobulin A (sIgA) levels in saliva were also measured. RESULTS: Consumption of AXOS-enriched breads led to increased faecal butyrate and a trend for reduced iso-valerate and fatty acids associated with protein fermentation. Faecal levels of bifidobacteria increased following initial control breads and remained elevated throughout the study. Lactobacilli levels were elevated following both placebo and AXOS-breads. No changes in salivary secretory IgA levels were observed during the study. Furthermore, no adverse effects on gastrointestinal symptoms were reported during AXOS-bread intake. CONCLUSIONS: AXOS-breads led to a potentially beneficial shift in fermentation end products and are well tolerated.

Adaptive noise cancellation with fast tunable RBF network

This paper describes a novel adaptive noise cancellation system with fast tunable radial basis function (RBF). The weight coefficients of the RBF network are adapted by the multi-innovation recursive least square (MRLS) algorithm. If the RBF network performs poorly despite of the weight adaptation, an insignificant node with little contribution to the overall performance is replaced with a new node without changing the model size. Otherwise, the RBF network structure remains unchanged and only the weight vector is adapted. The simulation results show that the proposed approach can well cancel the noise in both stationary and nonstationary ANC systems.

Online modeling with tunable RBF network

In this paper, we propose a novel online modeling algorithm for nonlinear and nonstationary systems using a radial basis function (RBF) neural network with a fixed number of hidden nodes. Each of the RBF basis functions has a tunable center vector and an adjustable diagonal covariance matrix. A multi-innovation recursive least square (MRLS) algorithm is applied to update the weights of RBF online, while the modeling performance is monitored. When the modeling residual of the RBF network becomes large in spite of the weight adaptation, a node identified as insignificant is replaced with a new node, for which the tunable center vector and diagonal covariance matrix are optimized using the quantum particle swarm optimization (QPSO) algorithm. The major contribution is to combine the MRLS weight adaptation and QPSO node structure optimization in an innovative way so that it can track well the local characteristic in the nonstationary system with a very sparse model. Simulation results show that the proposed algorithm has significantly better performance than existing approaches.

Candidate genes for obesity-susceptibility show enriched association within a large genome-wide association study for BMI

Before the advent of genome-wide association studies (GWASs), hundreds of candidate genes for obesity-susceptibility had been identified through a variety of approaches. We examined whether those obesity candidate genes are enriched for associations with body mass index (BMI) compared with non-candidate genes by using data from a large-scale GWAS. A thorough literature search identified 547 candidate genes for obesity-susceptibility based on evidence from animal studies, Mendelian syndromes, linkage studies, genetic association studies and expression studies. Genomic regions were defined to include the genes ±10 kb of flanking sequence around candidate and non-candidate genes. We used summary statistics publicly available from the discovery stage of the genome-wide meta-analysis for BMI performed by the genetic investigation of anthropometric traits consortium in 123 564 individuals. Hypergeometric, rank tail-strength and gene-set enrichment analysis tests were used to test for the enrichment of association in candidate compared with non-candidate genes. The hypergeometric test of enrichment was not significant at the 5% P-value quantile (P = 0.35), but was nominally significant at the 25% quantile (P = 0.015). The rank tail-strength and gene-set enrichment tests were nominally significant for the full set of genes and borderline significant for the subset without SNPs at P < 10(-7). Taken together, the observed evidence for enrichment suggests that the candidate gene approach retains some value. However, the degree of enrichment is small despite the extensive number of candidate genes and the large sample size. Studies that focus on candidate genes have only slightly increased chances of detecting associations, and are likely to miss many true effects in non-candidate genes, at least for obesity-related traits.

The consumer acceptance of novel vegetable-enriched bread products as a potential vehicle to increase vegetable consumption

Currently UK fruit and vegetable intakes are below recommendations. Bread is a staple food consumed by ~95% of adults in western countries. In addition, bread provides an ideal matrix by which functionality can be delivered to the consumer in an accepted food. Therefore, enriching bread with vegetables may be an effective strategy to increase vegetable consumption. This study evaluated consumer acceptance, purchase intent and intention of product replacement of bread enriched with red beetroot, carrot with coriander, red pepper with tomato or white beetroot (80g vegetable per serving of 200g) compared to white control bread (0g vegetable). Consumers (n=120) rated their liking of the breads overall, as well as their liking of appearance, flavour and texture using nine-point hedonic scales. Product replacement and purchase intent of the breads was rated using five-point scales. The effect of providing consumers with health information about the breads was also evaluated. There were significant differences in overall liking (P<0.0001), as well as liking of appearance (P<0.0001), flavour (P=0.0002) and texture (P=0.04), between the breads. However, the significant differences resulted from the red beetroot bread which was significantly (P<0.05) less liked compared to control bread. There were no significant differences in overall liking between any of the other vegetable-enriched breads compared with the control bread (no vegetable inclusion), apart from the red beetroot bread which was significantly less liked. The provision of health information about the breads did not increase consumer liking of the vegetable-enriched breads. In conclusion, this study demonstrated that vegetable-enriched bread appeared to be an acceptable strategy to increase vegetable intake, however, liking depended on vegetable type.

Rhodosporidium toruloides cultivated in NaCl-enriched glucose-based media: adaptation dynamics and lipid production

In the present report and for the first time in the international literature, the impact of the addition of NaCl upon growth and lipid production on the oleaginous yeast Rhodosporidium toruloides was studied. Moreover, equally for first time, lipid production by R. toruloides was performed under non-aseptic conditions. Therefore, the potentiality of R. toruloides DSM 4444 to produce lipid in media containing several initial concentrations of NaCl with glucose employed as carbon source was studied. Preliminary batch-flask trials with increasing amounts of NaCl revealed the tolerance of the strain against NaCl content up to 6.0% (w/v). However, 4.0% (w/v) of NaCl stimulated lipid accumulation for this strain, by enhancing lipid production up to 71.3% (w/w) per dry cell weight. The same amount of NaCl was employed in pasteurized batch-flask cultures in order to investigate the role of the salt as bacterial inhibiting agent. The combination of NaCl and high glucose concentrations was found to satisfactorily suppress bacterial contamination of R. toruloides cultures under these conditions. Batch-bioreactor trials of the yeast in the same media with high glucose content (up to 150 g/L) resulted in satisfactory substrate assimilation, with almost linear kinetic profile for lipid production, regardless of the initial glucose concentration imposed. Finally, fed-batch bioreactor cultures led to the production of 37.2 g/L of biomass, accompanied by 64.5% (w/w) of lipid yield. Lipid yield per unit of glucose consumed received the very satisfactory value of 0.21 g/g, a value amongst the highest ones in the literature. The yeast lipid produced contained mainly oleic acid and to lesser extent palmitic and stearic acids, thus constituting a perfect starting material for “second generation” biodiesel

A fast adaptive tunable RBF network for nonstationary systems

This paper describes a novel on-line learning approach for radial basis function (RBF) neural network. Based on an RBF network with individually tunable nodes and a fixed small model size, the weight vector is adjusted using the multi-innovation recursive least square algorithm on-line. When the residual error of the RBF network becomes large despite of the weight adaptation, an insignificant node with little contribution to the overall system is replaced by a new node. Structural parameters of the new node are optimized by proposed fast algorithms in order to significantly improve the modeling performance. The proposed scheme describes a novel, flexible, and fast way for on-line system identification problems. Simulation results show that the proposed approach can significantly outperform existing ones for nonstationary systems in particular.