To develop an end-to-end deep learning method for automated quantitative assessment of pediatric blunt hepatic trauma based on contrast-enhanced computed tomography (CT). This retrospective study included 170 children with blunt hepatic trauma between May 1, 2015, and August 30, 2021, who had undergone contrast-enhanced CT. Both liver parenchyma and liver trauma regions were manually segmented from CT images. Two deep convolutional neural networks (CNNs) were trained on 118 cases between May 1, 2015, and December 31, 2019, for liver segmentation and liver trauma segmentation. Liver volume and trauma volume were automatically calculated based on the segmentation results, and the liver parenchymal disruption index (LPDI) was computed as the ratio of liver trauma volume to liver volume. The segmentation performance was tested on 52 cases between January 1, 2020, and August 30, 2021. Correlation analysis among the LPDI, trauma volume, and the American Association for the Surgery of Trauma (AAST) liver injury grade was performed using the Spearman rank correlation. The performance of severity assessment of pediatric blunt hepatic trauma based on the LPDI and trauma volume was evaluated using receiver operating characteristic (ROC) analysis. The Dice, precision, and recall of the developed deep learning framework were 94.75, 94.11, and 95.46% in segmenting the liver and 72.91, 72.40, and 76.80% in segmenting the trauma regions. The LPDI and trauma volume were significantly correlated with AAST grade (rho = 0.823 and rho = 0.831, respectively; p < 0.001 for both). The area under the ROC curve (AUC) values for the LPDI and trauma volume to distinguish between high-grade and low-grade pediatric blunt hepatic trauma were 0.942 (95% CI, 0.882–1.000) and 0.952 (95% CI, 0.895–1.000), respectively. The developed end-to-end deep learning method is able to automatically and accurately segment the liver and trauma regions from contrast-enhanced CT images. The automated LDPI and liver trauma volume can act as objective and quantitative indexes to supplement the current AAST grading of pediatric blunt hepatic trauma.

Multiple imputation (MI) is a popular method for dealing with missing values. However, the suitable way for applying clustering after MI remains unclear: how to pool partitions? How to assess the clustering instability when data are incomplete? By answering both questions, this paper proposed a complete view of clustering with missing data using MI. The problem of partitions pooling is here addressed using consensus clustering while, based on the bootstrap theory, we explain how to assess the instability related to observed and missing data. The new rules for pooling partitions and instability assessment are theoretically argued and extensively studied by simulation. Partitions pooling improves accuracy while measuring instability with missing data enlarges the data analysis possibilities: it allows assessment of the dependence of the clustering to the imputation model, as well as a convenient way for choosing the number of clusters when data are incomplete, as illustrated on a real data set.

In recent years, probabilistic forecasting is an emerging topic, which is why there is a growing need of suitable methods for the evaluation of multivariate predictions. We analyze the sensitivity of the most common scoring rules, especially regarding quality of the forecasted dependency structures. Additionally, we propose scoring rules based on the copula, which uniquely describes the dependency structure for every probability distribution with continuous marginal distributions. Efficient estimation of the considered scoring rules and evaluation methods such as the Diebold-Mariano test are discussed. In detailed simulation studies, we compare the performance of the renowned scoring rules and the ones we propose. Besides extended synthetic studies based on recently published results we also consider a real data example. We find that the energy score, which is probably the most widely used multivariate scoring rule, performs comparably well in detecting forecast errors, also regarding dependencies. This contradicts other studies. The results also show that a proposed copula score provides very strong distinction between models with correct and incorrect dependency structure. We close with a comprehensive discussion on the proposed methodology.

Recovering sparse conditional independence graphs from data is a fundamental problem in machine learning with wide applications. A popular formulation of the problem is an $\ell_1$ regularized maximum likelihood estimation. Many convex optimization algorithms have been designed to solve this formulation to recover the graph structure. Recently, there is a surge of interest to learn algorithms directly based on data, and in this case, learn to map empirical covariance to the sparse precision matrix. However, it is a challenging task in this case, since the symmetric positive definiteness (SPD) and sparsity of the matrix are not easy to enforce in learned algorithms, and a direct mapping from data to precision matrix may contain many parameters. We propose a deep learning architecture, GLAD, which uses an Alternating Minimization (AM) algorithm as our model inductive bias, and learns the model parameters via supervised learning. We show that GLAD learns a very compact and effective model for recovering sparse graph from data.

Attention mechanisms have seen wide adoption in neural NLP models. In addition to improving predictive performance, these are often touted as affording transparency: models equipped with attention provide a distribution over attended-to input units, and this is often presented (at least implicitly) as communicating the relative importance of inputs. However, it is unclear what relationship exists between attention weights and model outputs. In this work, we perform extensive experiments across a variety of NLP tasks that aim to assess the degree to which attention weights provide meaningful `explanations' for predictions. We find that they largely do not. For example, learned attention weights are frequently uncorrelated with gradient-based measures of feature importance, and one can identify very different attention distributions that nonetheless yield equivalent predictions. Our findings show that standard attention modules do not provide meaningful explanations and should not be treated as though they do. Code for all experiments is available at https://github.com/successar/AttentionExplanation.

Modern large-scale machine learning applications require stochastic optimization algorithms to be implemented on distributed computational architectures. A key bottleneck is the communication overhead for exchanging information such as stochastic gradients among different workers. In this paper, to reduce the communication cost, we propose a convex optimization formulation to minimize the coding length of stochastic gradients. The key idea is to randomly drop out coordinates of the stochastic gradient vectors and amplify the remaining coordinates appropriately to ensure the sparsified gradient to be unbiased. To solve the optimal sparsification efficiently, several simple and fast algorithms are proposed for an approximate solution, with a theoretical guarantee for sparseness. Experiments on $\ell_2$ regularized logistic regression, support vector machines, and convolutional neural networks validate our sparsification approaches.

Many machine learning tasks are interested in recognizing or identifying an individual instance within a large set of possible candidates. These problems are usually modeled as multi-class classification problems, with a large and possibly complex label set. Leading examples include detecting the speaker from his voice patterns (Togneri and Pullella, 2011), identifying the author from her written text (Stamatatos et al., 2014), or labeling the object category from its image (Duygulu et al., 2002, Deng et al., 2010, Oquab et al., 2014). In all these examples, the algorithm observes an input x, and uses the classifier function h to guess the label y from a large label set S. 1 There are multiple practical challenges in developing classifiers for large label sets. Collecting high quality training data is perhaps the main obstacle, as the costs scale with the number of classes. It can be affordable to first collect data for a small set of classes, even if the long-term goal is to generalize to a larger set. Furthermore, classifier development can be accelerated by training first on fewer classes, as each training cycle may require substantially less resources. Indeed, due to interest in how small-set performance generalizes to larger sets, such comparisons can found in the literature (Oquab et al., 2014, Griffin et al., 2007). A natural question is: how does changing the size of the label set affect the classification accuracy?

Modern large scale machine learning applications require scaling stochastic optimization algorithms to distributed computational architectures. A key bottleneck is the communication overhead for exchanging information among different workers. For example, we have n training data distributed on M workers, and each of them owns its local copy of the model parameter vector. In the synchronized stochastic gradient method, each worker processes a random minibatch of its training data, and then the local updates are synchronized by making an All-Reduce step, which aggregates stochastic gradients from all workers, and taking a Broadcast step that transmits the updated parameter vector back to all workers. The process is repeated until an appropriate convergence criterion is met. An important factor that may significantly slow down any optimization algorithm is the communication cost among workers.

Variable selection for optimal treatment regime in a clinical trial or an observational study is getting more attention. Most existing variable selection techniques focused on selecting variables that are important for prediction, therefore some variables that are poor in prediction but are critical for decision-making may be ignored. A qualitative interaction of a variable with treatment arises when treatment effect changes direction as the value of this variable varies. The qualitative interaction indicates the importance of this variable for decision-making. Gunter et al. (2011) proposed S-score which characterizes the magnitude of qualitative interaction of each variable with treatment individually. In this article, we developed a sequential advantage selection method based on the modified S-score. Our method selects qualitatively interacted variables sequentially, and hence excludes marginally important but jointly unimportant variables {or vice versa}. The optimal treatment regime based on variables selected via joint model is more comprehensive and reliable. With the proposed stopping criteria, our method can handle a large amount of covariates even if sample size is small. Simulation results show our method performs well in practical settings. We further applied our method to data from a clinical trial for depression.