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.

FDTD analysis of close-coupled 418 MHz radiating devices for human biotelemetry

This paper describes the finite-difference time-domain (FDTD) analysis of antenna-body interaction effects occurring when chest-mounted 418 MHz radio transmitters are used for medical telemetry applications. Whole-body software models (homogeneous, layered and tissue-segmented) were developed for an adult male subject. Using an electrically small (300 mm(2)) planar loop antenna, calculated radiation efficiencies ranged between 33.5% and 39.2% for a whole-body model, and between 60.7% and 66.1% for a torso; radiation patterns were found to be largely independent of model composition. The computed radiation efficiency for a 21.5 kg phantom representing a six-year-old female was within 1.1 dB of measured results (actual body mass 28 kg) and well-correlated azimuthal radiation patterns were noted.

Organizational Change and Employees' Mental Health: The Protective Role of Sense of Coherence

Objective: To examine the impact of sense of coherence (SOC) on psychiatric events in the context of organizational merger. Methods: Data were derived from a prospective

Innate immunogenicity and in vitro protective potential of Schistosoma mansoni lung schistosomula excretory-secretory candidate vaccine antigens

Excretory secretory products (ESP) of Schistosoma mansoni developing larvae are ideal potential vaccines as such molecules may readily induce host primary immune responses, and local memory immune response effectors that would target, surround, and pursue the larvae while negotiating the lung blood capillaries. We herein characterized the cytokines response ESP, e.g., SG3PDH, 14-3-3-like protein, TPX, and calpain induce in the natural context of infection, and defined the global cytokine profile conducive to effective schistosome larvae killing. Accordingly, spleen cells (SC) taken from naive, and 7-, or 9-day S. mansoni-infected mice were stimulated in vitro with the selected ESP, in a recombinant or multiple antigen peptide (MAP) form, and examined for production of T helper type (Th) 1, Th2, and Th17 cytokines, and the ability to mediate in vitro attrition of lung-stage schistosomula. The study indicated that larval ESP principally elicit Th1 and Th17 type cytokines. Recombinant SG3PDH was the only test ESP to additionally activate SC from S. mansoni-infected BALB/c mice to release higher IL-4 levels than unstimulated SC and mediate significant (P

Synthesis of protective Mo-Si-B coatings in molten salts and their oxidation behavior in an air-water mixture

The formation of various phases during boronizing of silicided molybdenum substrates (MoSi2/Mo) was investigated. Boronizing treatments were conducted in molten salts under an inert gas atmosphere in the 700-1000 degrees C temperature range for 3-7 h. Depending on the process type (non-current or electrochemical) and molten salt temperature, the formation of different boride phases (MoB, Mo2B5, MoB2, MoB4) was observed. At the same time, substantial oxidation of the bulk molybdenum disilicide phase (MoSi2) to the Mo5Si3 phase was observed in non-current boronizing. The oxidation resistance of the coatings was investigated by the weight change in an air-water (2.3 vol.%) mixture at a temperature of 500 degrees C for a period up to 700 h. Results indicated that a two-phase microstructure consisting of the MoSi2, matrix phase with 12-15 wt.% of the MoB4 phase greatly improved the oxidation resistance of the molybdenum substrates. The weight gain rate observed was 6.5 center dot 10(-4) mg/cm(2) h. (c) 2006 Elsevier B.V. All rights reserved.