Beyond Expectation: Deep Joint Mean and Quantile Regression for Spatiotemporal Problems.

Spatiotemporal problems are ubiquitous and of vital importance in many research fields. Despite the potential already demonstrated by deep learning methods in modeling spatiotemporal data, typical approaches tend to focus solely on conditional expectations of the output variables being modeled. In this article, we propose a multioutput multiquantile deep learning approach for jointly modeling several conditional quantiles together with the conditional expectation as a way to provide a more complete "picture" of the predictive density in spatiotemporal problems. Using two large-scale data sets from the transportation domain, we empirically demonstrate that, by approaching the quantile regression problem from a multitask learning perspective, it is possible to solve the embarrassing quantile crossings problem while simultaneously significantly outperforming state-of-the-art quantile regression methods. Moreover, we show that jointly modeling the mean and several conditional quantiles not only provides a rich description about the predictive density that can capture heteroscedastic properties at a neglectable computational overhead but also leads to improved predictions of the conditional expectation due to the extra information and the regularization effect induced by the added quantiles.

A Bayesian Additive Model for Understanding Public Transport Usage in Special Events.

Public special events, like sports games, concerts and festivals are well known to create disruptions in transportation systems, often catching the operators by surprise. Although these are usually planned well in advance, their impact is difficult to predict, even when organisers and transportation operators coordinate. The problem highly increases when several events happen concurrently. To solve these problems, costly processes, heavily reliant on manual search and personal experience, are usual practice in large cities like Singapore, London or Tokyo. This paper presents a Bayesian additive model with Gaussian process components that combines smart card records from public transport with context information about events that is continuously mined from the Web. We develop an efficient approximate inference algorithm using expectation propagation, which allows us to predict the total number of public transportation trips to the special event areas, thereby contributing to a more adaptive transportation system. Furthermore, for multiple concurrent event scenarios, the proposed algorithm is able to disaggregate gross trip counts into their most likely components related to specific events and routine behavior. Using real data from Singapore, we show that the presented model outperforms the best baseline model by up to 26 percent in R2 and also has explanatory power for its individual components.

Learning Supervised Topic Models for Classification and Regression from Crowds.

The growing need to analyze large collections of documents has led to great developments in topic modeling. Since documents are frequently associated with other related variables, such as labels or ratings, much interest has been placed on supervised topic models. However, the nature of most annotation tasks, prone to ambiguity and noise, often with high volumes of documents, deem learning under a single-annotator assumption unrealistic or unpractical for most real-world applications. In this article, we propose two supervised topic models, one for classification and another for regression problems, which account for the heterogeneity and biases among different annotators that are encountered in practice when learning from crowds. We develop an efficient stochastic variational inference algorithm that is able to scale to very large datasets, and we empirically demonstrate the advantages of the proposed model over state-of-the-art approaches.

Competing risks mixture model for traffic incident duration prediction.

Traffic incident duration is known to result from a combination of multiple factors, including covariates such as spatial and temporal characteristics, traffic conditions, and existence of secondary accidents but also the clearance method itself. In this paper, a competing risks mixture model is used to investigate the influence of clearance methods and various covariates on the duration of traffic incidents and predict traffic incident duration. The proposed mixture model considers the uncertainty in any of five clearance methods that occurred. The probability of the clearance method is specified in the mixture by using a multinomial logistic model. Three candidate distributions, namely, generalized gamma, Weibull, and log-logistic are tested to determine the most appropriate probability density function of the parametric survival analysis model. The unobserved heterogeneity is also incorporated into the mixture model in a way that allows parameters to vary across observations based on the three candidate distributions. The methods are illustrated with incident data from Singaporean expressways from January 2010 to December 2011. Regression analysis reveals that the probability of different clearance methods and the duration of traffic incidents are both significantly affected by various factors, such as traffic conditions and incident characteristics. Results show that the proposed mixture model is better than the traditional accelerated failure time model, and it predicts traffic incident duration with reasonable accuracy, as shown by the mean average percent error.

Transplantation of Schwann cells and/or olfactory ensheathing glia into the contused spinal cord: Survival, migration, axon association, and functional recovery.

Schwann cells (SCs) and olfactory ensheathing glia (OEG) have shown promise for spinal cord injury repair. We sought their in vivo identification following transplantation into the contused adult rat spinal cord at 1 week post-injury by: (i) DNA in situ hybridization (ISH) with a Y-chromosome specific probe to identify male transplants in female rats and (ii) lentiviral vector-mediated expression of EGFP. Survival, migration, and axon-glia association were quantified from 3 days to 9 weeks post-transplantation. At 3 weeks after transplantation into the lesion, a 60-90% loss of grafted cells was observed. OEG-only grafts survived very poorly within the lesion (<5%); injection outside the lesion led to a 60% survival rate, implying that the injury milieu was hostile to transplanted cells and or prevented their proliferation. At later times post-grafting, p75(+)/EGFP(-) cells in the lesion outnumbered EGFP(+) cells in all paradigms, evidence of significant host SC infiltration. SCs and OEG injected into the injury failed to migrate from the lesion. Injection of OEG outside of the injury resulted in their migration into the SC-injected injury site, not via normal-appearing host tissue but along the pia or via the central canal. In all paradigms, host axons were seen in association with or ensheathed by transplanted glia. Numerous myelinated axons were found within regions of grafted SCs but not OEG. The current study details the temporal survival, migration, axon association of SCs and OEG, and functional recovery after grafting into the contused spinal cord, research previously complicated due to a lack of quality, long-term markers for cell tracking in vivo.

Gold nanoelectrode ensembles for direct trace electroanalysis of iodide.

A procedure for the standardization of ensembles of gold nanodisk electrodes (NEE) of 30 nm diameter is presented, which is based on the analytical comparison between experimental cyclic voltammograms (CV) obtained at the NEEs in diluted solutions of redox probes and CV patterns obtained by digital simulation. Possible origins of defects sometimes found in NEEs are discussed. Selected NEEs are then employed for the study of the electrochemical oxidation of iodide in acidic solutions. CV patterns display typical quasi-reversible behavior which involves associated chemical reactions between adsorbed and solution species. The main CV characteristics at the NEE compare with those observed at millimeter sized gold disk electrodes (Au-macro), apart a slight shift in E1/2 values and slightly higher peak to peak separation at the NEE. The detection limit (DL) at NEEs is 0.3 microM, which is more than one order of magnitude lower than DL at the Au-macro (4 microM). The mechanism of the electrochemical oxidation of iodide at NEEs is discussed. Finally, NEEs are applied to the direct determination of iodide at micromolar concentration levels in real samples, namely in some ophthalmic drugs and iodized table salt.

Inhibition of tumour necrosis factor-alpha by antisense targeting produces immunophenotypical and morphological changes in injury-activated microglia and macrophages.

Microglia respond in a stereotypical pattern to a diverse array of pathological states. These changes are coupled to morphological and immunophenotypical alterations and the release of a variety of reactive species, trophic factors and cytokines that modify both microglia and their cellular environment. We examined whether a microglial-produced cytokine, tumour necrosis factor-alpha (TNF-alpha), was involved in the maintenance of microglial activation after spinal cord injury by selective inhibition using TNF-alpha antisense deoxyoligonucleotides (ASOs). Microglia and macrophages harvested from 3 d post-contused rat spinal cord were large and rounded (86.3 +/- 9.6%). They were GSA-IB4-positive (GSA-IB4(+)) (Griffonia simplicifolia lectin, microglia specific; 94.8 +/- 5.1%), strongly OX-42 positive (raised against a type 3 complement/integrin receptor, CD11b; 78.9 +/- 9.1%), ED-1 positive (a lysosomal marker shown to correlate well with immune cell activation; 97.2 +/- 2.6%) and IIA positive (antibody recognizes major histocompatibility complex II; 57.2 +/- 5.6%), indicative of fully activated cells, for up to 48 h after plating. These cells also secreted significant amounts of TNF-alpha (up to 436 pg/microg total protein, 16 h). Fluoroscein isothiocyanate-labelled TNF-alpha ASOs (5, 50 and 200 nm) added to the culture medium were taken up very efficiently into the cells (> 90% cells) and significantly reduced TNF-alpha production by up to 92% (26.5 pg/microg total protein, 16 h, 200 nm TNF-alpha ASOs). Furthermore, few of the treated cells at this time were round (5.4 +/- 2.7%), having become predominantly spindle shaped (74.9 +/- 6.3%) or stellate (21.4 +/- 2.7%); immunophenotypically, although all of them remained GSA-IB4 positive (91.6 +/- 6.2%), many were weakly OX-42 positive and few expressed either ED-1 (12.9 +/- 2.5%) or IIA (19.8 +/- 7.4%). Thus, the secretion of TNF-alpha early in spinal cord injury may be involved in autoactivating microglia/macrophages. However, at the peak of microglial activation after injury, the activation state of microglia/macrophages is not stable and this process may still be reversible by blocking TNF-alpha.

cAMP and Schwann cells promote axonal growth and functional recovery after spinal cord injury.

Central neurons regenerate axons if a permissive environment is provided; after spinal cord injury, however, inhibitory molecules are present that make the local environment nonpermissive. A promising new strategy for inducing neurons to overcome inhibitory signals is to activate cAMP signaling. Here we show that cAMP levels fall in the rostral spinal cord, sensorimotor cortex and brainstem after spinal cord contusion. Inhibition of cAMP hydrolysis by the phosphodiesterase IV inhibitor rolipram prevents this decrease and when combined with Schwann cell grafts promotes significant supraspinal and proprioceptive axon sparing and myelination. Furthermore, combining rolipram with an injection of db-cAMP near the graft not only prevents the drop in cAMP levels but increases them above those in uninjured controls. This further enhances axonal sparing and myelination, promotes growth of serotonergic fibers into and beyond grafts, and significantly improves locomotion. These findings show that cAMP levels are key for protection, growth and myelination of injured CNS axons in vivo and recovery of function.

Voltammetric characteristics of miconazole and its cathodic stripping voltammetric determination.

Miconazole is reduced at mercury electrode above pH 6 involving organometallic compound formation, responsible for an anomalous polarographic behavior. The electrodic process presents a large contribution of the adsorption effects. The drug can be determined by cathodic stripping voltammetry from 8.0 x 10(-8) to 1, 5 x 10(-6) molL-1 in Britton-Robinson buffer pH 8.0, when pre-accumulated for 30s at an accumulation potential of 0V. A relative standard deviation of 3.8% was obtained for ten measurements of 1.0 x 10(-7) molL-1 miconazole in B-R buffer pH 8.0 and a limit detection of 1, 7 x 10(-8) molL-1 was determined using 60s of deposition time and scan rate of 100 mVs-1. The proposed method is simple, precise and it was applied successfully for the determination of the miconazole in pure form and in commercial formulations, showing mean recoveries of 99.7-98.4%.

Voltammetric characteristics of miconazole and its cathodic stripping voltammetric determination

Miconazole is reduced at mercury electrode above pH 6 involving organometallic compound formation, responsible for an anomalous polarographic behavior. The electrodic process presents a large contribution of the adsorption effects. The drug can be determined by cathodic stripping voltammetry from 8.0 x 10-8 to 1, 5 x 10-6 molL-1 in Britton-Robinson buffer pH 8.0, when pre-accumulated for 30s at an accumulation potential of 0V. A relative standard deviation of 3.8 percent was obtained for ten measurements of 1.0 x 10-7 molL-1 miconazole in B-R buffer pH 8.0 and a limit detection of 1, 7 x 10-8 molL-1 was determined using 60s of deposition time and scan rate of 100 mVs-1. The proposed method is simple, precise and it was applied successfully for the determination of the miconazole in pure form and in commercial formulations, showing mean recoveries of 99.7-98.4 percent

The Use of Antisense-Mediated Inhibition to Delineate The Role of Inflammatory Agents in The Pathophysiology of Spinal Cord Injury.

Injuries to the central nervous system (CNS) usually lead to a potent and acute inflammatory response[1]. During this period, glia and immune cells respond to chemical cues associated with the debris of lysed neurons, disrupted axons, and a broken blood-brain-barrier by releasing a battery of cytokines including tumor necrosis factor-α (TNF-α) and, interleukin-ß (IL-1ß) as well as reactive oxygen species such as nitric oxide (NO-)[2]. The secretion of these factors may be primarily responsible for secondary damage to surrounding uninjured tissue that potentiates the initial injury[3]. Antisense oligonucleotides (ASOs) are designed to hybridize to specific regions of specific mRNAs. Hybridization of the oligonucleotide to the mRNA then interferes with the normal processing of that mRNA at the ribosome or targets the RNA duplex for cleavage by the RNA digestive enzyme, ribonuclease H, resulting in greatly reduced expression of the coded protein. This effectively reduces the amount of corresponding translated protein product and experiments can be designed to examine the requirement of particular inflammatory agents in eliciting specific deleterious responses after injury, e.g., cell death.