Explore the vast range of titles available.

This research proposes constructing a network used for person re-identification called MGNACP (Multiple Granularity Network with Attention Mechanisms and Combination Poolings). Based on the MGN (Multiple Granularity Network) that combines global and local features and the characteristics of the MGN branch, the MGNA (Multiple Granularity Network with Attentions) is designed by adding a channel attention mechanism to each global and local branch of the MGN. The MGNA, with attention mechanisms, learns the most identifiable information about global and local features to improve the person re-identification accuracy. Based on the constructed MGNA, a single pooling used in each branch is replaced by combination pooling to form MGNACP. The combination pooling parameters are the proportions of max pooling and average pooling in combination pooling. Through experiments, suitable combination pooling parameters are found, the advantages of max pooling and average pooling are preserved and enhanced, and the disadvantages of both types of pooling are overcome, so that poolings can achieve optimal results in MGNACP and improve the person re-identification accuracy. In experiments on the Market-1501 dataset, MGNACP achieved competitive experimental results; the values of mAP and top-1 are 88.82% and 95.46%. The experimental results demonstrate that MGNACP is a competitive person re-identification network, and that the attention mechanisms and combination poolings can significantly improve the person re-identification accuracy.

Background Missing data prove troublesome in data analysis; at best they reduce a study’s statistical power and at worst they induce bias in parameter estimates. Multiple imputation via chained equations is a popular technique for dealing with missing data. However, techniques for combining and pooling results from fitted generalized additive models (GAMs) after multiple imputation have not been well explored. Methods We simulated missing data under MCAR, MAR, and MNAR frameworks and utilized random forest and predictive mean matching imputation to investigate a variety of rules for combining GAMs after multiple imputation with binary and normally distributed outcomes. We compared multiple pooling procedures including the “D2” method, the Cauchy combination test, and the median p -value (MPV) rule. The MPV rule involves simply computing and reporting the median p -value across all imputations. Other ad hoc methods such as a mean p -value rule and a single imputation method are investigated. The viability of these methods in pooling results from B-splines is also examined for normal outcomes. An application of these various pooling techniques is then performed on two case studies, one which examines the effect of elevation on a six-minute walk distance (a normal outcome) for patients with pulmonary arterial hypertension, and the other which examines risk factors for intubation in hospitalized COVID-19 patients (a dichotomous outcome). Results In comparison to the results from generalized additive models fit on full datasets, the median p -value rule performs as well as if not better than the other methods examined. In situations where the alternative hypothesis is true, the Cauchy combination test appears overpowered and alternative methods appear underpowered, while the median p -value rule yields results similar to those from analyses of complete data. Conclusions For pooling results after fitting GAMs to multiply imputed datasets, the median p -value is a simple yet useful approach which balances both power to detect important associations and control of Type I errors.

We consider two ways to aggregate expert opinions using simple averages: averaging probabilities and averaging quantiles. We examine analytical properties of these forecasts and compare their ability to harness the wisdom of the crowd. In terms of location, the two average forecasts have the same mean. The average quantile forecast is always sharper: it has lower variance than the average probability forecast. Even when the average probability forecast is overconfident, the shape of the average quantile forecast still offers the possibility of a better forecast. Using probability forecasts for gross domestic product growth and inflation from the Survey of Professional Forecasters, we present evidence that both when the average probability forecast is overconfident and when it is underconfident, it is outperformed by the average quantile forecast. Our results show that averaging quantiles is a viable alternative and indicate some conditions under which it is likely to be more useful than averaging probabilities. This paper was accepted by Peter Wakker, decision analysis.

To identify statistical methods for harmonization, the procedures aimed at achieving the comparability of previously collected data, which could be used in the context of summary data and individual participant data meta-analysis of cognitive measures. Environmental scan methods were used to conduct two reviews to identify (1) studies that quantitatively combined data on cognition and (2) general literature on statistical methods for data harmonization. Search results were rapidly screened to identify articles of relevance. All 33 meta-analyses combining cognition measures either restricted their analyses to a subset of studies using a common measure or combined standardized effect sizes across studies; none reported their harmonization steps before producing summary effects. In the second scan, three general classes of statistical harmonization models were identified (1) standardization methods, (2) latent variable models, and (3) multiple imputation models; few publications compared methods. Although it is an implicit part of conducting a meta-analysis or pooled analysis, the methods used to assess inferential equivalence of complex constructs are rarely reported or discussed. Progress in this area will be supported by guidelines for the conduct and reporting of the data harmonization and integration and by evaluating and developing statistical approaches to harmonization.

The evaluation of hedge fund performance is challenging given the flexible nature of hedge funds' strategies and their lack of operational transparency. As a result, inference about skill is inevitably contaminated by the error in the benchmark model. To address this concern, we propose a model pooling approach to develop a fund-specific benchmark obtained by pooling a set of diverse attribution models. The weights assigned to the individual models in the pool are based on the log score criterion, an information-theoretic measure of the conditional performance of a model. We illustrate the advantages of a pooled benchmark over alternative approaches, including the Fung and Hsieh [Fung W, Hsieh DA (2004) Hedge fund benchmarks: A risk-based approach. Financial Analysts J. 60: 65-80] model, stepwise regression methods, and style-adjusted methods in the contexts of a real-time investment strategy, hedge fund replication, and fund failure prediction.

Hand-crafted and learning-based features are two main types of video representations in the field of video understanding. How to integrate their merits to design good descriptors has been the research hotspot recently. Motivated by TDD (Wang et al. 2015), we combine trajectory pooling method and 3D ConvNets (Tran et al. 2015) and put forward a novel multi-scale trajectory-pooled 3D convolutional descriptor (MTC3D) for action recognition in this paper. Specifically, we calculate multi-scale dense trajectories from the input video and perform trajectory pooling on feature maps of 3D CNN. The proposed descriptor has two advantages: 3D CNN has the ability to extract high-level semantic information from videos and multi-scale trajectory pooling method utilizes the temporal information of videos subtly. The experiments on the datasets of HMDB51 and UCF101 demonstrate that the proposed descriptor achieves state-of-the-art results.

BackgroundThis article describes a new procedure to estimate the mean and variance of greenhouse gases (GHG) emission factors based on different, possibly conflicting, estimates for these emission factors. The procedure uses common information such as mean and standard deviation usually reported in IPCC (Intergovernmental Panel on Climate Change) database and other references in the literature that estimate emission factors. Essentially, it is a procedure in the class of meta-analysis, based on the computation of Sa2, a new estimator for the variance of the emission factor.ResultsWe discuss the quality of this estimator in terms of its probability distribution and show that it is unbiased. The resulting confidence interval for the mean emission factor is tighter than those that would have resulted from using other estimators such as pooled variance and thus, the new procedure improves the accuracy in estimating GHG emissions.The application of the procedure is illustrated in a case study involving the estimation of methane emissions from rice cultivation.ConclusionsThe estimation of emission factors using Sa2 was demonstrated to be more accurate because it is not biased and more precise than alternative methods.

In this paper, we propose to extend the flexibility of the commonly used 2 × 2 non-overlapping max pooling for Convolutional Neural Network. We name it as Bi-linearly Weighted Fractional Max-Pooling. This proposed method enables max pooling operation below stride size 2, and is computed based on four bi-linearly weighted neighboring input activations. Currently, in a 2 × 2 non-overlapping max pooling operation, as spatial size is halved in both x and y directions, three-quarter of activations in the feature maps are discarded. As such reduction is too abrupt, amount of said pooling operation within a Convolutional Neural Network is very limited: further increasing the number of pooling operation results in too little activation left for subsequent operations. Using our proposed pooling method, spatial size reduction can be more gradual and can be adjusted flexibly. We applied a few combinations of our proposed pooling method into 50-layered ResNet and 19-layered VGGNet with reduced number of filters, and experimented on FGVC-Aircraft, Oxford-IIIT Pet, STL-10 and CIFAR-100 datasets. Even with reduced memory usage, our proposed methods showed reasonable improvement in classification accuracy with 50-layered ResNet. Additionally, with flexibility of our proposed pooling method, we change the reduction rate dynamically every training iteration, and our evaluation results indicated potential regularization effect.

We prove merge theorems along the entire sequence of natural numbers for the distribution functions of suitably centered and normed linear combinations of independent and identically distributed random variables from the domain of geometric partial attraction of any non-normal semistable law. Surprisingly, for some sequences of linear combinations, not too far from those with equal weights, the merge theorems reduce to ordinary asymptotic distributions with semistable limits. The proofs require working out general conditions for merging in terms of characteristic functions.

Merging asymptotic expansions are established for the distribution functions of suitably centered and normed linear combinations of winnings in a full sequence of generalized St. Petersburg games, where a linear combination is viewed as the share of any one of n cooperative gamblers who play with a pooling strategy. The expansions are given in terms of Fourier-Stieltjes transforms and are constructed from suitably chosen members of the classes of subsequential semistable infinitely divisible asymptotic distributions for the total winnings of the n players and from their pooling strategy, where the classes themselves are determined by the two parameters of the game. For all values of the tail parameter, the expansions yield best possible rates of uniform merge. Surprisingly, it turns out that for a subclass of strategies, not containing the averaging uniform strategy, our merging approximations reduce to asymptotic expansions of the usual type, derived from a proper limiting distribution. The Fourier-Stieltjes transforms are shown to be numerically invertible in general and it is also demonstrated that the merging expansions provide excellent approximations even for very small n.