Counting craters in remotely sensed images is the only tool that provides relative dating of remote planetary surfaces. Surveying craters requires counting a large amount of small subkilometer craters, which calls for highly efficient automatic crater detection. In this article, we present an integrated framework on autodetection of subkilometer craters with boosting and transfer learning. The framework contains three key components. First, we utilize mathematical morphology to efficiently identify crater candidates, the regions of an image that can potentially contain craters. Only those regions occupying relatively small portions of the original image are the subjects of further processing. Second, we extract and select image texture features, in combination with supervised boosting ensemble learning algorithms, to accurately classify crater candidates into craters and noncraters. Third, we integrate transfer learning into boosting, to enhance detection performance in the regions where surface morphology differs from what is characterized by the training set. Our framework is evaluated on a large test image of 37, 500 × 56, 250 m 2 on Mars, which exhibits a heavily cratered Martian terrain characterized by nonuniform surface morphology. Empirical studies demonstrate that the proposed crater detection framework can achieve an F1 score above 0.85, a significant improvement over the other crater detection algorithms.

Weak supervision has shown promising results in many natural language processing tasks, such as Named Entity Recognition (NER). Existing work mainly focuses on learning deep NER models only with weak supervision, i.e., without any human annotation, and shows that by merely using weakly labeled data, one can achieve good performance, though still underperforms fully supervised NER with manually/strongly labeled data. In this paper, we consider a more practical scenario, where we have both a small amount of strongly labeled data and a large amount of weakly labeled data. Unfortunately, we observe that weakly labeled data does not necessarily improve, or even deteriorate the model performance (due to the extensive noise in the weak labels) when we train deep NER models over a simple or weighted combination of the strongly labeled and weakly labeled data. To address this issue, we propose a new multi-stage computational framework -NEEDLE with three essential ingredients: (1) weak label completion, (2) noise-aware loss function, and (3) final finetuning over the strongly labeled data. Through experiments on E-commerce query NER and Biomedical NER, we demonstrate that NEE-DLE can effectively suppress the noise of the weak labels and outperforms existing methods. In particular, we achieve new SOTA F1-scores on 3 Biomedical NER datasets: BC5CDRchem 93.74, BC5CDR-disease 90.69, NCBIdisease 92.28.

Abstract-Instead of assuming fully loaded cells in the analysis on cache-enabled networks with tools of stochastic geometry, we focus on the dynamic traffic in this letter. With modeling traffic dynamics of request arrivals and departures, probabilities of full-, free-, and modest-load cells in the large-scale cache-enabled network are elaborated based on the traffic queue state. Moreover, we propose to exploit the packets cached at cache-enabled users as side information to cancel the incoming interference. Then the packet loss rates for both the cache-enabled and cache-untenable users are investigated. The simulation results verify our analysis.

Both increased arterial stiffness and higher total homocysteine (tHcy) are associated with an elevated risk for cardiovascular disease. However, the relationship between tHcy and arterial stiffness is still inconclusive. The authors aimed to test the relationship of tHcy with carotid-femoral pulse wave velocity (cfPWV) and examine the possible effect modifiers in adults. A study was conducted from July to September 2016 in Jiangsu Province, China. A total of 16 644 participants were enrolled in the final analysis. Increased arterial stiffness is defined as a cfPWV ≥10 m/s. Overall, there was a positive association between tHcy and cfPWV levels (per 5-μmol/L tHcy increase: β = 0.10; 95% confidence interval [CI], 0.08-0.13) and increased arterial stiffness (per 5-μmol/L tHcy increase: odds ratio, 1.11; 95% CI, 1.07-1.14). Compared with participants with tHcy <10 μmol/L, the significantly higher cfPWV levels were observed in those with tHcy ≥15 μmol/L (β = 0.37; 95% CI, 0.28-0.47). Accordingly, a higher prevalence of increased arterial stiffness was found in patients with tHcy10 to <15 μmol/L (odds ratio, 1.18; 95% CI, 1.05-1.33) and tHcy ≥15 μmol/L (odds ratio, 1.50; 95% CI, 1.32-1.71) as compared with participants with tHcy <10 μmol/L. Furthermore, the stronger positive association was found in participants who were older (≥60 years, P for interaction = .008), had low body mass index (<25 kg/m , P for interaction = .026), high systolic blood pressure levels (≥145 mm Hg [median], P for interaction = .048), or diabetes mellitus (P for interaction = .045). The present study demonstrated that serum tHcy concentrations were positively associated with cfPWV and the prevalence of increased arterial stiffness. These results suggest that the cardiovascular effects of tHcy may partly be mediated through arterial stiffness.

Influence maximization in a social network is to target a given number of nodes in the network such that the expected number of activated nodes from these nodes is maximized. A social network usually exhibits some degree of modularity. Previous research efforts that made use of this topological property are restricted to random networks with two communities. In this paper, we firstly transform the influence maximization problem in a modular social network to an optimal resource allocation problem in the same network. We assume that the communities of the social network are disconnected. We then propose a recursive relation for finding such an optimal allocation. We prove that finding an optimal allocation in a modular social network is NP-hard and propose a new optimal dynamic programming algorithm to solve the problem. We name our new algorithm OASNET(Optimal Allocation in a Social NETwork). We compare OASNET with equal allocation, proportional allocation, random allocation and selecting top degree nodes without any allocation strategy on both synthetic and real world datasets. Experimental results show that OASNET outperforms these four heuristics.