An efficient feature selection method for mobile devices with application to activity recognition

This paper presents a feature selection method for data classification, which combines a model-based variable selection technique and a fast two-stage subset selection algorithm. The relationship between a specified (and complete) set of candidate features and the class label is modelled using a non-linear full regression model which is linear-in-the-parameters. The performance of a sub-model measured by the sum of the squared-errors (SSE) is used to score the informativeness of the subset of features involved in the sub-model. The two-stage subset selection algorithm approaches a solution sub-model with the SSE being locally minimized. The features involved in the solution sub-model are selected as inputs to support vector machines (SVMs) for classification. The memory requirement of this algorithm is independent of the number of training patterns. This property makes this method suitable for applications executed in mobile devices where physical RAM memory is very limited. An application was developed for activity recognition, which implements the proposed feature selection algorithm and an SVM training procedure. Experiments are carried out with the application running on a PDA for human activity recognition using accelerometer data. A comparison with an information gain based feature selection method demonstrates the effectiveness and efficiency of the proposed algorithm.

An Investigation into the Dissolution Properties of Celecoxib Melt Extrudates: Understanding the Role of Polymer Type and Concentration in Stabilizing Supersaturated Drug Concentrations

In this study, the dissolution properties of celecoxib (CX) solid dispersions manufactured from Eudragit 4155F and polyvinylpyrrolidone (PVP) were evaluated. Hot-melt extrusion (HME) technology was used to prepare amorphous solid dispersions of drug/polymer binary systems at different mass ratios. The drug concentrations achieved from the dissolution of PVP and Eudragit 4155F solid dispersions in phosphate buffer, pH 7.4 (PBS 7.4) were significantly greater than the equilibrium solubility of CX (1.58 µg/mL). The degree of supersaturation increased significantly as the polymer concentration within the solid dispersion increased. The maximum drug concentration achieved by PVP solid dispersions did not significantly exceed the apparent solubility of amorphous CX. The predominant mechanism for achieving supersaturated CX concentrations in PBS 7.4 was attributed to stabilization of amorphous CX during dissolution. Conversely, Eudragit 4155F solid dispersions showed significantly greater supersaturated drug solutions particularly at high polymer concentrations. For example, at a drug/polymer ratio of 1:9, a concentration of 100 µg/mL was achieved after 60 min that was stable (no evidence of drug recrystallization) for up to 72 h. This clearly identifies the potential of Eudragit 4155F to act as a solubilizing agent for CX. These findings were in good agreement with the results from solubility performed using PBS 7.4 in which Eudragit 4155F had been predissolved. In these tests, Eudragit 4155F significantly increased the equilibrium solubility of CX. Solution 1H NMR spectra were used to identify drug/polymer interactions. Deshielding of CX aromatic protons (H-1a and H-1b) containing the sulfonamide group occurred as a result of dissolution of Eudragit 4155F solid dispersions, whereas deshielding of H-1a protons and shielding of H-1b protons occurred as a result of the dissolution of PVP solid dispersions. In principle, it is reasonable to suggest that the different drug/polymer interactions observed give rise to the variation in dissolution observed for the two polymer/drug systems.

Model Selection in SVMs Using Differential Evolution

To improve the performance of classification using Support Vector Machines (SVMs) while reducing the model selection time, this paper introduces Differential Evolution, a heuristic method for model selection in two-class SVMs with a RBF kernel. The model selection method and related tuning algorithm are both presented. Experimental results from application to a selection of benchmark datasets for SVMs show that this method can produce an optimized classification in less time and with higher accuracy than a classical grid search. Comparison with a Particle Swarm Optimization (PSO) based alternative is also included.

In pursuit of parrhesia: an investigation into knowledge selection in inquisitorial planning practice

Drawing upon interviews with procedural actants from Public Inquiry and Examination in Public fora, I draw upon relevant theoretical frameworks to evaluate modes of discourse in inquisitorial planning practice. In the investigation, which is based primarily upon an empirical study, I focus upon the role of evidence, the selection and handling of multiple knowledges, the behaviour of participants, and the methodology underpinning the process. It is established that such arenas can be effective mechanisms for testing complex evidence; and suggestions are made for improved practice, procedure, and future research. I conclude by raising serious ethical questions concerning participant behaviour, particularly on the part of advocates and especially chartered town planners.

Comparison of techniques to screen and characterize bacteria-specific hybridomas for high-quality monoclonal antibodies selection

Antibodies are are very important materials for diagnostics. A rapid and simple hybridoma screening method will help in delivering specific monoclonal antibodies. In this study, we systematically developed the first antibody array to screen for bacteria-specific monoclonal antibodies using Listeria monocytogenes as a bacteria model. The antibody array was developed to expedite the hybridoma screening process by printing hybridoma supernatants on a glass slide coated with an antigen of interest. This screening method is based on the binding ability of supernatants to the coated antigen. The bound supernatants were detected by a fluorescently labeled anti-mouse immunoglobulin. Conditions (slide types, coating, spotting, and blocking buffers) for antibody array construction were optimized. To demonstrate its usefulness, antibody array was used to screen a sample set of 96 hybridoma supernatants in comparison to ELISA. Most of the positive results identified by ELISA and antibody array methods were in agreement except for those with low signals that were undetectable by antibody array. Hybridoma supernatants were further characterized with surface plasmon resonance to obtain additional data on the characteristics of each selected clone. While the antibody array was slightly less sensitive than ELISA, a much faster and lower cost procedure to screen clones against multiple antigens has been demonstrated. (C) 2011 Elsevier Inc. All rights reserved.