Polymer-carbon black composite sensors in an electronic nose for air-quality monitoring

An electronic nose that uses an array of 32 polymer-carbon black composite sensors has been developed, trained, and tested. By selecting a variety of chemical functionalities in the polymers used to make sensors, it is possible to construct an array capable of identifying and quantifying a broad range of target compounds, such as alcohols and aromatics, and distinguishing isomers and enantiomers (mirror-image isomers). A model of the interaction between target molecules and the polymer-carbon black composite sensors is under development to aid in selecting the array members and to enable identification of compounds with responses not stored in the analysis library.     

Improvements to the SMO algorithm for SVM regression

This paper points out an important source of inefficiency in Smola and Scholkopf's (1998) sequential minimal optimization (SMO) algorithm for support vector machine regression that is caused by the use of a single threshold value. Using clues from the Karush-Kuhn-Tucker conditions for the dual problem, two threshold parameters are employed to derive modifications of SMO for regression. These modified algorithms perform significantly faster than the original SMO on the datasets tried.     

A fast quasi-Newton method for semi-supervised SVM

Due to its wide applicability, semi-supervised learning is an attractive method for using unlabeled data in classification. In this work, we present a semi-supervised support vector classifier that is designed using quasi-Newton method for nonsmooth convex functions. The proposed algorithm is suitable in dealing with very large number of examples and features. Numerical experiments on various benchmark datasets showed that the proposed algorithm is fast and gives improved generalization performance over the existing methods. Further, a non-linear semi-supervised SVM has been proposed based on a multiple label switching scheme. This non-linear semi-supervised SVM is found to converge faster and it is found to improve generalization performance on several benchmark datasets.     

Fast generalized cross-validation algorithm for sparse model learning

We propose a fast, incremental algorithm for designing linear regression models. The proposed algorithm generates a sparse model by optimizing multiple smoothing parameters using the generalized cross-validation approach. The performances on synthetic and real-world data sets are compared with other incremental algorithms such as Tipping and Faul's fast relevance vector machine, Chen et al.'s orthogonal least squares, and Orr's regularized forward selection. The results demonstrate that the proposed algorithm is competitive.     

A Fast Dual Algorithm for Kernel Logistic Regression

This paper gives a new iterative algorithm for kernel logistic regression. It is based on the solution of a dual problem using ideas similar to those of the Sequential Minimal Optimization algorithm for Support Vector Machines. Asymptotic convergence of the algorithm is proved. Computational experiments show that the algorithm is robust and fast. The algorithmic ideas can also be used to give a fast dual algorithm for solving the optimization problem arising in the inner loop of Gaussian Process classifiers. Editor: Shai Ben-David     

Monitoring space shuttle air quality using the Jet Propulsion Laboratory electronic nose

A miniature electronic nose (ENose) has been designed and built at the Jet Propulsion Laboratory (JPL), Pasadena, CA, and was designed to detect, identify, and quantify ten common contaminants and relative humidity changes. The sensing array includes 32 sensing films made from polymer carbon-black composites. Event identification and quantification were done using the Levenberg-Marquart nonlinear least squares method. After successful ground training, this ENose was used in a demonstration experiment aboard STS-95 (October-November, 1998), in which the ENose was operated continuously for six days and recorded the sensors' response to the air in the mid-deck. Air samples were collected daily and analyzed independently after the flight. Changes in shuttle-cabin humidity were detected and quantified by the JPL ENose; neither the ENose nor the air samples detected any of the contaminants on the target list. The device is microgravity insensitive.     

Gas-phase hydroxylation of benzene over H–Ga-FER zeolite

The direct hydroxylation of benzene with N₂O is studied using H–Ga-FER zeolite as catalyst. The H–Ga-FER zeolite was synthesized by using the hydrothermal synthesis route. Pyrrolidine was used as a template. The catalyst was characterized by using XRD and NMR techniques. The gas-phase hydroxylation reaction was carried out using nitrous oxide gas as an oxidant. The reaction was thoroughly studied by changing different parameters like temperature, time, reactant mole ratio and contact time (WHSV). It was observed that by using H–Ga-FER, phenol was formed with 100% selectivity. The deactivation is faster in case of H–Ga-FER as compared with Al-FER and Fe-FER. This revised version was published online in July 2006 with corrections to the Cover Date.     

Heat transfer in rectangular microchannels during volumetric heating of the substrate

Convective heat transfer in microchannels with rectangular and square cross sections are analyzed for volumetric heat generation in the substrate due to an imposed magnetic field. Gadolinium was used as the substrate material and water as the working fluid. Gadolinium is a magnetic material that exhibits high temperature rise during adiabatic magnetization around its transition temperature of 295 K. A thorough investigation for velocity and temperature distributions was performed by varying channel aspect ratio, Reynolds number, and heat generation rate in the substrate. With the increase in Reynolds number, the outlet temperature decreased and the average Nusselt number increased.