An orthogonal forward regression technique for sparse kernel density estimation

Using the classical Parzen window (PW) estimate as the desired response, the kernel density estimation is formulated as a regression problem and the orthogonal forward regression technique is adopted to construct sparse kernel density (SKD) estimates. The proposed algorithm incrementally minimises a leave-one-out test score to select a sparse kernel model, and a local regularisation method is incorporated into the density construction process to further enforce sparsity. The kernel weights of the selected sparse model are finally updated using the multiplicative nonnegative quadratic programming algorithm, which ensures the nonnegative and unity constraints for the kernel weights and has the desired ability to reduce the model size further. Except for the kernel width, the proposed method has no other parameters that need tuning, and the user is not required to specify any additional criterion to terminate the density construction procedure. Several examples demonstrate the ability of this simple regression-based approach to effectively construct a SKID estimate with comparable accuracy to that of the full-sample optimised PW density estimate. (c) 2007 Elsevier B.V. All rights reserved.

Identification of nonlinear systems using generalized kernel models

Nonlinear system identification is considered using a generalized kernel regression model. Unlike the standard kernel model, which employs a fixed common variance for all the kernel regressors, each kernel regressor in the generalized kernel model has an individually tuned diagonal covariance matrix that is determined by maximizing the correlation between the training data and the regressor using a repeated guided random search based on boosting optimization. An efficient construction algorithm based on orthogonal forward regression with leave-one-out (LOO) test statistic and local regularization (LR) is then used to select a parsimonious generalized kernel regression model from the resulting full regression matrix. The proposed modeling algorithm is fully automatic and the user is not required to specify any criterion to terminate the construction procedure. Experimental results involving two real data sets demonstrate the effectiveness of the proposed nonlinear system identification approach.

Kernel classifier construction using orthogonal forward selection and boosting with Fisher ratio class separability measure

A greedy technique is proposed to construct parsimonious kernel classifiers using the orthogonal forward selection method and boosting based on Fisher ratio for class separability measure. Unlike most kernel classification methods, which restrict kernel means to the training input data and use a fixed common variance for all the kernel terms, the proposed technique can tune both the mean vector and diagonal covariance matrix of individual kernel by incrementally maximizing Fisher ratio for class separability measure. An efficient weighted optimization method is developed based on boosting to append kernels one by one in an orthogonal forward selection procedure. Experimental results obtained using this construction technique demonstrate that it offers a viable alternative to the existing state-of-the-art kernel modeling methods for constructing sparse Gaussian radial basis function network classifiers. that generalize well.

Fast kernel classifier construction using orthogonal forward selection to minimise leave-one-out misclassification rate

We propose a simple yet computationally efficient construction algorithm for two-class kernel classifiers. In order to optimise classifier's generalisation capability, an orthogonal forward selection procedure is used to select kernels one by one by minimising the leave-one-out (LOO) misclassification rate directly. It is shown that the computation of the LOO misclassification rate is very efficient owing to orthogonalisation. Examples are used to demonstrate that the proposed algorithm is a viable alternative to construct sparse two-class kernel classifiers in terms of performance and computational efficiency.

A kernel-based two-class classifier for imbalanced data sets

Many kernel classifier construction algorithms adopt classification accuracy as performance metrics in model evaluation. Moreover, equal weighting is often applied to each data sample in parameter estimation. These modeling practices often become problematic if the data sets are imbalanced. We present a kernel classifier construction algorithm using orthogonal forward selection (OFS) in order to optimize the model generalization for imbalanced two-class data sets. This kernel classifier identification algorithm is based on a new regularized orthogonal weighted least squares (ROWLS) estimator and the model selection criterion of maximal leave-one-out area under curve (LOO-AUC) of the receiver operating characteristics (ROCs). It is shown that, owing to the orthogonalization procedure, the LOO-AUC can be calculated via an analytic formula based on the new regularized orthogonal weighted least squares parameter estimator, without actually splitting the estimation data set. The proposed algorithm can achieve minimal computational expense via a set of forward recursive updating formula in searching model terms with maximal incremental LOO-AUC value. Numerical examples are used to demonstrate the efficacy of the algorithm.

A forward-constrained regression algorithm for sparse kernel density estimation

Using the classical Parzen window (PW) estimate as the target function, the sparse kernel density estimator is constructed in a forward-constrained regression (FCR) manner. The proposed algorithm selects significant kernels one at a time, while the leave-one-out (LOO) test score is minimized subject to a simple positivity constraint in each forward stage. The model parameter estimation in each forward stage is simply the solution of jackknife parameter estimator for a single parameter, subject to the same positivity constraint check. For each selected kernels, the associated kernel width is updated via the Gauss-Newton method with the model parameter estimate fixed. The proposed approach is simple to implement and the associated computational cost is very low. Numerical examples are employed to demonstrate the efficacy of the proposed approach.

Production of haploids and doubled haploids in oil palm

Background: Oil palm is the world’s most productive oil-food crop despite yielding well below its theoretical maximum. This maximum could be approached with the introduction of elite F1 varieties. The development of such elite lines has thus far been prevented by difficulties in generating homozygous parental types for F1 generation. Results: Here we present the first high-throughput screen to identify spontaneously-formed haploid (H) and doubled haploid (DH) palms. We secured over 1,000 Hs and one DH from genetically diverse material and derived further DH/mixoploid palms from Hs using colchicine. We demonstrated viability of pollen from H plants and expect to generate 100% homogeneous F1 seed from intercrosses between DH/mixoploids once they develop female inflorescences. Conclusions: This study has generated genetically diverse H/DH palms from which parental clones can be selected in sufficient numbers to enable the commercial-scale breeding of F1 varieties. The anticipated step increase in productivity may help to relieve pressure to extend palm cultivation, and limit further expansion into biodiverse rainforest.

Long chain n-3 PUFA-rich meal reduced postprandial measures of arterial stiffness

Background & aims The consumption of long chain n − 3 polyunsaturated fatty acids (LC n − 3 PUFA) is known to be cardio-protective. Data on the influence of LC n − 3 PUFA on arterial stiffness in the postprandial state is limited. The aim of this study was to investigate the acute effects of a LC n − 3 PUFA-rich meal on measures of arterial stiffness. Methods Twenty-five healthy subjects (12 men, 13 women) received a control and a LC n − 3 PUFA-rich meal on two occasions in a random order. Arterial stiffness was measured at baseline, 30, 60, 90, 120, 180 and 240 min after meal consumption by pulse wave analysis and digital volume pulse to derive an augmentation index and a stiffness index respectively. Blood samples were taken for measurement of lipids, glucose and insulin. Results Consumption of the LC n − 3 PUFA-rich meal had an attenuating effect on augmentation index (P = 0.02) and stiffness index (P = 0.03) compared with the control meal. A significant treatment effect (P = 0.036) was seen for plasma non-esterified fatty acids concentrations. Conclusions These data indicate that acute LC n − 3 PUFA-rich meal consumption can improve postprandial arterial stiffness. This has important implications for the beneficial properties of LC n − 3 PUFA and cardiovascular risk reduction.

Sparse kernel density estimation technique based on zero-norm constraint

A sparse kernel density estimator is derived based on the zero-norm constraint, in which the zero-norm of the kernel weights is incorporated to enhance model sparsity. The classical Parzen window estimate is adopted as the desired response for density estimation, and an approximate function of the zero-norm is used for achieving mathemtical tractability and algorithmic efficiency. Under the mild condition of the positive definite design matrix, the kernel weights of the proposed density estimator based on the zero-norm approximation can be obtained using the multiplicative nonnegative quadratic programming algorithm. Using the -optimality based selection algorithm as the preprocessing to select a small significant subset design matrix, the proposed zero-norm based approach offers an effective means for constructing very sparse kernel density estimates with excellent generalisation performance.

Regression based D-optimality experimental design for sparse kernel density estimation

This paper derives an efficient algorithm for constructing sparse kernel density (SKD) estimates. The algorithm first selects a very small subset of significant kernels using an orthogonal forward regression (OFR) procedure based on the D-optimality experimental design criterion. The weights of the resulting sparse kernel model are then calculated using a modified multiplicative nonnegative quadratic programming algorithm. Unlike most of the SKD estimators, the proposed D-optimality regression approach is an unsupervised construction algorithm and it does not require an empirical desired response for the kernel selection task. The strength of the D-optimality OFR is owing to the fact that the algorithm automatically selects a small subset of the most significant kernels related to the largest eigenvalues of the kernel design matrix, which counts for the most energy of the kernel training data, and this also guarantees the most accurate kernel weight estimate. The proposed method is also computationally attractive, in comparison with many existing SKD construction algorithms. Extensive numerical investigation demonstrates the ability of this regression-based approach to efficiently construct a very sparse kernel density estimate with excellent test accuracy, and our results show that the proposed method compares favourably with other existing sparse methods, in terms of test accuracy, model sparsity and complexity, for constructing kernel density estimates.

Acute effects of meal fatty acid composition on insulin sensitivity in healthy post-menopausal women

Postprandial plasma insulin concentrations after a single high-fat meal may be modified by the presence of specific fatty acids although the effects of sequential meal ingestion are unknown. The aim of the present study was to examine the effects of altering the fatty acid composition in a single mixed fat-carbohydrate meal on glucose metabolism and insulin sensitivity of a second meal eaten 5 h later. Insulin sensitivity was assessed using a minimal model approach. Ten healthy post-menopausal women underwent four two-meal studies in random order. A high-fat breakfast (40 g fat) where the fatty acid composition was predominantly saturated fatty acids (SFA), n-6 polyunsaturated fatty acids (PUFA), long-chain n-3 PUFA or monounsaturated fatty acids (MUFA) was followed 5 h later by a low-fat, high-carbohydrate lunch (5.7 g fat), which was identical in all four studies. The plasma insulin response was significantly higher following the SFA meal than the other meals after both breakfast and lunch (P<0.006) although there was no effect of breakfast fatty acid composition on plasma glucose concentrations. Postprandial insulin sensitivity (SI(Oral)) was assessed for 180 min after each meal. SI(Oral) was significantly lower after lunch than after breakfast for all four test meals (P=0.019) following the same rank order (SFA < n-6 PUFA < n-3 PUFA < MUFA) for each meal. The present study demonstrates that a single meal rich in SFA reduces postprandial insulin sensitivity with 'carry-over' effects for the next meal.

Measurement of apolipoprotein B-48 in the Svedberg flotation rate (Sf) > 400, Sf 60 – 400 and Sf 20 – 60 lipoprotein fractions reveals novel findings with respect to the effects of dietary fatty acids on triacylglycerol-rich lipoproteins in postmenopausal women

The present study was designed to examine whether the type of fat ingested in an initial test meal influences the response and density distribution of dietary-derived lipoproteins in the Svedberg flotation rate (Sf)>400, Sf 60 - 400 and Sf 20 - 60 lipoprotein fractions. A single-blind randomized within-subject crossover design was used to study the effects of palm oil, safflower oil, a mixture of fish and safflower oil, and olive oil on postprandial apolipoprotein (apo) B-48, retinyl ester and triacylglycerol responses in each lipoprotein fraction following an initial test meal containing one of the oils and a second standardized test meal. For all dietary oils, late postprandial (300min) concentrations of triacylglycerol and apo B-48 were significantly higher in the Sf 60 - 400 fraction than in the Sf>400 fraction (P<0.02). Significantly greater apo B-48 incremental areas under the curve (IAUCs) were also observed in the Sf 60 - 400 fraction than in the Sf>400 fraction following palm oil, safflower oil and olive oil (P<0.04), with a similar non-significant trend for fish/safflower oil. Olive oil resulted in a significantly greater apo B-48 IAUC in the Sf>400 fraction (P<0.02) than did any of the other dietary oils, as well as a tendency for a higher IAUC in the Sf 60 - 400 fraction compared with the palm, safflower and fish/safflower oils. In conclusion, we have found that the majority of intestinally derived lipoproteins present in the circulation following meals enriched with saturated, polyunsaturated or monounsaturated fatty acids are of the density and size of small chylomicrons and chylomicron remnants. Olive oil resulted in a greater apo B-48 response compared with the other dietary oils following sequential test meals, suggesting the formation of a greater number of small (Sf 60 - 400) and large (Sf>400) apo B-48-containing lipoproteins in response to this dietary oil.

Acute ingestion of a meal rich in n-3 polyunsaturated fatty acids results in rapid gastric emptying in humans

Background: n-3 Polyunsaturated fatty acids (PUFAs) have proven benefits for both the development of atherosclerosis and inflammatory conditions. The effects on atherosclerosis may be partly mediated by the observed reduction in fasting and postprandial triacylglycerol concentrations after both acute and chronic n-3 PUFA ingestion. Objective: The aim of this study was to assess gastric emptying and gastrointestinal hormone release after the consumption of mixed meals rich in n-3 PUFAs or other classes of fatty acids. Design: Ten healthy women (aged 50–62 y) completed 4 separate study visits in a single-blind, randomized design. On each occasion, subjects consumed 40 g oil rich in either saturated fatty acids, monounsaturated fatty acids, n-6 PUFAs, or n-3 PUFAs as part of a mixed meal. [1-13C]Octanoic acid (100 mg) was added to each oil. Gastric emptying was assessed by a labeled octanoic acid breath test, and concentrations of gastrointestinal hormones and plasma lipids were measured. Results: Recovery of 13C in breath was enhanced after n-3 PUFA ingestion (P < 0.005). The cholecystokinin response after the n-3 PUFA meal was significantly delayed (P < 0.001), and the glucagon-like peptide 1 response was significantly reduced (P < 0.05). Conclusion: The inclusion of n-3 PUFAs in a meal alters the gastric emptying rate, potentially as the result of changes in the pattern of cholecystokinin and glucagon-like peptide 1 release.