Trap spectroscopy by the glow rate technique using bleaching of colour centres

An application of the glow rate technique (GRT) for analysis of the parameters of thermostimulated decay of colour centres is presented using the data on the decay of radiation defects in LiBaF3:Fe crystals created by X rays at 300 K. The GRT offers a procedure for evaluation of the mean activation energy as a function of temperature in the case of arbitrary thermostimulated relaxation kinetics represented by the trap distribution function. The experimental procedure involves at least two subsequent measurements of thermostimulated decay kinetics at different heating rates. It is shown that the decay of the F type centres is governed by interaction of mobile anion vacancies with F(A) and F centres, leading to both the hopping migration and recombination of F centres and the thermoactivated dissociation of the F(A) centres.     

Interactions of Ta2O5 with water vapor at elevated temperatures

Tantalum pentoxide (Ta₂O₅) and its solid solution phases are candidate coatings for components to be used in combustion environments. Thus, it is important to understand the response of Ta₂O₅ to high‐temperature water vapor, a product of combustion. Thermogravimetric methods are used to examine the oxide in reactant streams of controlled water vapor contents at 1250°C‐1450°C. The observed weight loss indicates a reaction of the general form ½ Ta₂O₅(s) + x H₂O(g)=TaO_y(OH)_x(g). Methodical variation in the water vapor pressure suggests the products are a mix of TaO(OH)₃(g) and Ta(OH)₅(g). Evidence of TaO(OH)₃(g) was observed with a sampling mass spectrometer. The measured hydroxide and oxyhyroxide vapor fluxes from Ta₂O₅ are compared with calculated vapor fluxes from SiO₂ and Al₂O₃. Ta₂O₅ exhibits fluxes similar to those from SiO₂ due to gaseous metal hydroxide formation.     

Fast Similarity Search for Learned Metrics

We introduce a method that enables scalable similarity search for learned metrics. Given pairwise similarity and dissimilarity constraints between some examples, we learn a Mahalanobis distance function that captures the examples' underlying relationships well. To allow sublinear time similarity search under the learned metric, we show how to encode the learned metric parameterization into randomized locality-sensitive hash functions. We further formulate an indirect solution that enables metric learning and hashing for vector spaces whose high dimensionality makes it infeasible to learn an explicit transformation over the feature dimensions. We demonstrate the approach applied to a variety of image data sets, as well as a systems data set. The learned metrics improve accuracy relative to commonly used metric baselines, while our hashing construction enables efficient indexing with learned distances and very large databases.     

Kernelized locality-sensitive hashing

Fast retrieval methods are critical for many large-scale and data-driven vision applications. Recent work has explored ways to embed high-dimensional features or complex distance functions into a low-dimensional Hamming space where items can be efficiently searched. However, existing methods do not apply for high-dimensional kernelized data when the underlying feature embedding for the kernel is unknown. We show how to generalize locality-sensitive hashing to accommodate arbitrary kernel functions, making it possible to preserve the algorithm's sublinear time similarity search guarantees for a wide class of useful similarity functions. Since a number of successful image-based kernels have unknown or incomputable embeddings, this is especially valuable for image retrieval tasks. We validate our technique on several data sets, and show that it enables accurate and fast performance for several vision problems, including example-based object classification, local feature matching, and content-based retrieval.     

Thin-Layer Gravity Current with Implications for Desalination Brine Disposal

Measurements of stratification and dissolved oxygen (DO) illustrate a hypersaline gravity current with salt loads similar to a desalination plant brine discharge. Over a 48-h sampling period in August 2005, alternating cycles of high- and low-temperature hypersaline water were observed along the bottom of Corpus Christi Bay in Texas, coincident with low benthic DO and tidal flushing from an adjacent smaller bay. The gravity current underflow was typically less than 10% of the overall water depth. Strong salinity gradients prevented wind-mixing of the entire water column. Hypoxic and near-hypoxic conditions were associated with limited DO replenishment from the ambient water. High DO levels in the underflow source water did not deter the development of offshore benthic hypoxia. A quasi-Lagrangian analysis is used to evaluate the relationship between ambient mixing and lateral mixing within the underflow. The analysis is further applied to estimating DO demand rates in the hypersaline plume. Mixing between the ambient water and the underflow predominately occurs over the sloping bay boundary. Once the gravity current reaches the flatter section of the bay, mixing is substantially reduced and DO is progressively depleted at the bottom. The transit time of the underflow (i.e., residence time or isolation time for water near the bottom) and wind-mixing energy appear to be key factors governing stratification persistence and potential hypoxia development. The observations and analyses provide insight into possible fate, impacts, and open questions associated with similarly scaled salt loadings from a desalination plant into a shallow bay.     

Semi-supervised graph clustering: a kernel approach

Semi-supervised clustering algorithms aim to improve clustering results using limited supervision. The supervision is generally given as pairwise constraints; such constraints are natural for graphs, yet most semi-supervised clustering algorithms are designed for data represented as vectors. In this paper, we unify vector-based and graph-based approaches. We first show that a recently-proposed objective function for semi-supervised clustering based on Hidden Markov Random Fields, with squared Euclidean distance and a certain class of constraint penalty functions, can be expressed as a special case of the weighted kernel k-means objective (Dhillon et al., in Proceedings of the 10th International Conference on Knowledge Discovery and Data Mining, 2004a). A recent theoretical connection between weighted kernel k-means and several graph clustering objectives enables us to perform semi-supervised clustering of data given either as vectors or as a graph. For graph data, this result leads to algorithms for optimizing several new semi-supervised graph clustering objectives. For vector data, the kernel approach also enables us to find clusters with non-linear boundaries in the input data space. Furthermore, we show that recent work on spectral learning (Kamvar et al., in Proceedings of the 17th International Joint Conference on Artificial Intelligence, 2003) may be viewed as a special case of our formulation. We empirically show that our algorithm is able to outperform current state-of-the-art semi-supervised algorithms on both vector-based and graph-based data sets.     

Asymmetric and Category Invariant Feature Transformations for Domain Adaptation

-1We address the problem of visual domain adaptation for transferring object models from one dataset or visual domain to another. We introduce a unified flexible model for both supervised and semi-supervised learning that allows us to learn transformations between domains. Additionally, we present two instantiations of the model, one for general feature adaptation/alignment, and one specifically designed for classification. First, we show how to extend metric learning methods for domain adaptation, allowing for learning metrics independent of the domain shift and the final classifier used. Furthermore, we go beyond classical metric learning by extending the method to asymmetric, category independent transformations. Our framework can adapt features even when the target domain does not have any labeled examples for some categories, and when the target and source features have different dimensions. Finally, we develop a joint learning framework for adaptive classifiers, which outperforms competing methods in terms of multi-class accuracy and scalability. We demonstrate the ability of our approach to adapt object recognition models under a variety of situations, such as differing imaging conditions, feature types, and codebooks. The experiments show its strong performance compared to previous approaches and its applicability to large-scale scenarios.     

Physician satisfaction in a changing health care environment: the impact of challenges to professional autonomy, authority, and dominance

Despite the widespread use of self-report measures of both job-related stressors and strains, relatively few carefully developed scales for which validity data exist are available. In this article, we discuss 3 job stressor scales (Interpersonal Conflict at Work Scale, Organizational Constraints Scale, and Quantitative Workload Inventory) and 1 job strain scale (Physical Symptoms Inventory). Using meta-analysis, we combined the results of 18 studies to provide estimates of relations between our scales and other variables. Data showed moderate convergent validity for the 3 job stressor scales, suggesting some objectively to these self-reports. Norms for each scale are provided.     

Weighted graph cuts without eigenvectors a multilevel approach

A variety of clustering algorithms have recently been proposed to handle data that is not linearly separable; spectral clustering and kernel k-means are two of the main methods. In this paper, we discuss an equivalence between the objective functions used in these seemingly different methods--in particular, a general weighted kernel k-means objective is mathematically equivalent to a weighted graph clustering objective. We exploit this equivalence to develop a fast, high-quality multilevel algorithm that directly optimizes various weighted graph clustering objectives, such as the popular ratio cut, normalized cut, and ratio association criteria. This eliminates the need for any eigenvector computation for graph clustering problems, which can be prohibitive for very large graphs. Previous multilevel graph partitioning methods, such as Metis, have suffered from the restriction of equal-sized clusters; our multilevel algorithm removes this restriction by using kernel k-means to optimize weighted graph cuts. Experimental results show that our multilevel algorithm outperforms a state-of-the-art spectral clustering algorithm in terms of speed, memory usage, and quality. We demonstrate that our algorithm is applicable to large-scale clustering tasks such as image segmentation, social network analysis and gene network analysis.     

Occurrence of antibiotics in hospital, residential, and dairy effluent, municipal wastewater, and the Rio Grande in New Mexico

This study had three objectives: 1) determine occurrence of antibiotics in effluent from hospitals, residential facilities, and dairies, and in municipal wastewater 2) determine antibiotic removal at a large wastewater treatment plant (WWTP) in Albuquerque, NM, and 3) determine concentrations of antibiotics in the Rio Grande, which receives wastewater from the Albuquerque WWTP. Twenty-three samples of wastewater and 3 samples of Rio Grande water were analyzed for the presence of 11 antibiotics. Fifty-eight percent of samples had at least one antibiotic present while 25% had three or more. Hospital effluent had detections of sulfamethoxazole, trimethoprim, ciprofloxacin, ofloxacin, lincomycin, and penicillin G, with 4 of 5 hospital samples having at least one antibiotic detected and 3 having four or more. At the residential sampling sites, ofloxacin was found in effluent from assisted living and retirement facilities, while the student dormitory had no detects. Only lincomycin was detected in dairy effluent (in 2 of 8 samples, at 700 and 6600 ng/L). Municipal wastewater had detections of sulfamethoxazole, trimethoprim, ciprofloxacin, and ofloxacin, with 4 of 6 samples having at least one antibiotic present and 3 having 3 or more. The relatively high concentrations (up to 35,500 ng/L) of ofloxacin found in hospital and residential effluent may be of concern due to potential genotoxic effects and development of antibiotic resistance. At the Albuquerque WWTP, both raw wastewater and treated effluent had detections of sulfamethoxazole, trimethoprim, and ofloxacin, at concentrations ranging from 110 to 470 ng/L. However, concentrations in treated effluent were reduced by 20% to 77%. No antibiotics were detected in the Rio Grande upstream of the Albuquerque WWTP discharge, and only one antibiotic, sulfamethoxazole, was detected in the Rio Grande (300 ng/L) below the WWTP.