Warped Mixtures for Nonparametric Cluster Shapes

A mixture of Gaussians fit to a single curved or heavy-tailed cluster will report that the data contains many clusters. To produce more appropriate clusterings, we introduce a model which warps a latent mixture of Gaussians to produce nonparametric cluster shapes. The possibly low-dimensional latent mixture model allows us to summarize the properties of the high-dimensional clusters (or density manifolds) describing the data. The number of manifolds, as well as the shape and dimension of each manifold is automatically inferred. We derive a simple inference scheme for this model which analytically integrates out both the mixture parameters and the warping function. We show that our model is effective for density estimation, performs better than infinite Gaussian mixture models at recovering the true number of clusters, and produces interpretable summaries of high-dimensional datasets.

Low molar mass organosiloxane liquid crystals for telecommunication applications

Mixtures of two proprietary low molar mass organosiloxane liquid crystals were studied in order to improve their alignment and optimize their electro-optic properties for telecommunication applications. Over a certain concentration range, mixtures exhibited an isotropic-chiral smectic A-chiral smectic C (Iso-SmA*-SmC*) phase sequence leading to exceptionally good alignment. At room temperature, the spontaneous polarization of these samples was reduced from 225 nC cm -2 in the pure SmC* liquid crystal to as low as 75 nC cm -2 in the mixture. Within this concentration range, the ferroelectric tilt angle could be varied between 35° and 15°, while the rise time decreased by 69.4%. The rise times were < 45 μs for moderate electric fields of ± 10 V μm -1 in the SmC* phase and ∼ 4 μs, independent of electric field, in the SmA* phase. At λ = 1550 nm, these mixtures exhibited very large extinction ratios of {\sim} 60 dB for binary switching in the SmC* phase and ∼ 55 dB continuous variable attenuation in the SmA* phase. © 2012 IEEE.

Impact of heterogeneous link qualities and network connectivity on binary consensus

In this paper we consider a network that is trying to reach consensus over the occurrence of an event while communicating over Additive White Gaussian Noise (AWGN) channels. We characterize the impact of different link qualities and network connectivity on consensus performance by analyzing both the asymptotic and transient behaviors. More specifically, we derive a tight approximation for the second largest eigenvalue of the probability transition matrix. We furthermore characterize the dynamics of each individual node. © 2009 AACC.

A 5-step reduced mechanism for combustion of CO/H2/H2O/CH4/CO2 mixtures with low hydrogen/methane and high H2O content

In this study a 5-step reduced chemical kinetic mechanism involving nine species is developed for combustion of Blast Furnace Gas (BFG), a multi-component fuel containing CO/H2/CH4/CO2, typically with low hydrogen, methane and high water fractions, for conditions relevant for stationary gas-turbine combustion. This reduced mechanism is obtained from a 49-reaction skeletal mechanism which is a modified subset of GRI Mech 3.0. The skeletal and reduced mechanisms are validated for laminar flame speeds, ignition delay times and flame structure with available experimental data, and using computational results with a comprehensive set of elementary reactions. Overall, both the skeletal and reduced mechanisms show a very good agreement over a wide range of pressure, reactant temperature and fuel mixture composition. © 2012 The Combustion Institute..

Model reductions for inference: generality of pairwise, binary, and planar factor graphs.

We offer a solution to the problem of efficiently translating algorithms between different types of discrete statistical model. We investigate the expressive power of three classes of model-those with binary variables, with pairwise factors, and with planar topology-as well as their four intersections. We formalize a notion of "simple reduction" for the problem of inferring marginal probabilities and consider whether it is possible to "simply reduce" marginal inference from general discrete factor graphs to factor graphs in each of these seven subclasses. We characterize the reducibility of each class, showing in particular that the class of binary pairwise factor graphs is able to simply reduce only positive models. We also exhibit a continuous "spectral reduction" based on polynomial interpolation, which overcomes this limitation. Experiments assess the performance of standard approximate inference algorithms on the outputs of our reductions.

Encoding of mixtures in a simple olfactory system.

Natural odors are usually mixtures; yet, humans and animals can experience them as unitary percepts. Olfaction also enables stimulus categorization and generalization. We studied how these computations are performed with the responses of 168 locust antennal lobe projection neurons (PNs) to varying mixtures of two monomolecular odors, and of 174 PNs and 209 mushroom body Kenyon cells (KCs) to mixtures of up to eight monomolecular odors. Single-PN responses showed strong hypoadditivity and population trajectories clustered by odor concentration and mixture similarity. KC responses were much sparser on average than those of PNs and often signaled the presence of single components in mixtures. Linear classifiers could read out the responses of both populations in single time bins to perform odor identification, categorization, and generalization. Our results suggest that odor representations in the mushroom body may result from competing optimization constraints to facilitate memorization (sparseness) while enabling identification, classification, and generalization.

Encoding of mixtures in a simple olfactory system

Natural odors are usually mixtures; yet, humans and animals can experience them as unitary percepts. Olfaction also enables stimulus categorization and generalization. We studied how these computations are performed with the responses of 168 locust antennal lobe projection neurons (PNs) to varying mixtures of two monomolecular odors, and of 174 PNs and 209 mushroom body Kenyon cells (KCs) to mixtures of up to eight monomolecular odors. Single-PN responses showed strong hypoadditivity and population trajectories clustered by odor concentration and mixture similarity. KC responses were much sparser on average than those of PNs and often signaled the presence of single components in mixtures. Linear classifiers could read out the responses of both populations in single time bins to perform odor identification, categorization, and generalization. Our results suggest that odor representations in the mushroom body may result from competing optimization constraints to facilitate memorization (sparseness) while enabling identification, classification, and generalization

Simulations of autoignition and laminar premixed flames in methane/air mixtures diluted with hot products

This article considers constant-pressure autoignition and freely propagating premixed flames of cold methane/air mixtures mixed with equilibrium hot products at high enough dilution levels to burn within the moderate to intense low oxygen dilution (MILD) combustion regime. The analysis is meant to provide further insight on MILD regime boundaries and to identify the effect of hot products speciation. As the mass fraction of hot products in the reactants mixture increases, autoignition occurs earlier. Species profiles show that the products/reactants mixture approximately equilibrates to a new state over a quick transient well before the main autoignition event, but as dilution becomes very high, this equilibration transient becomes more prominent and eventually merges with the primary ignition event. The dilution level at which these two reactive zones merge corresponds well with that marking the transition into the MILD regime, as defined according to conventional criteria. Similarly, premixed flame simulations at high dilutions show evidence of significant reactions involving intermediate species prior to the flame front. Since the premixed flame governing equations system demands that the species and temperature gradients be zero at the "cold" boundary, flame speed cannot be calculated above a certain dilution level. Up to this point, which again agrees reasonably well with the transition into the MILD regime according to convention, the laminar burning velocity was found to increase with hot product dilution while flame thickness remained largely unchanged. Some comments on the MILD combustion regime boundary definition for gas turbine applications are included. Copyright © Taylor & Francis Group, LLC.

Femtosecond laser induced breakdown spectroscopy of air-methane mixtures

Femtosecond laser pulses are used in order to induce dielectric breakdown in gaseous mixtures, namely in some reactive air-methane mixtures. The light emitted from the laser induced plasma was analyzed while the main emission features are identified and assigned. From the analysis of the emission spectra, a linear relationship was found to hold between the intensity of some spectral features and methane content. Finally, the use of femtosecond laser induced breakdown as a tool for the in situ determination of the composition of gaseous mixtures (e.g., equivalence ratio) is also discussed. © 2013 Elsevier B.V. All rights reserved.