Online linear programming plays an important role in both revenue management and resource allocation, and recent research has focused on developing efficient first-order online learning algorithms. Despite the empirical success of first-order methods, they typically achieve a regret no better than $\mathcal{O}(\sqrt{T})$, which is suboptimal compared to the $\mathcal{O}(\log T)$ bound guaranteed by the state-of-the-art linear programming (LP)-based online algorithms. This paper establishes several important facts about online linear programming, which unveils the challenge for first-order-method-based online algorithms to achieve beyond $\mathcal{O}(\sqrt{T})$ regret. To address the challenge, we introduce a new algorithmic framework that decouples learning from decision-making. More importantly, for the first time, we show that first-order methods can attain regret $\mathcal{O}(T^{1/3})$ with this new framework. Lastly, we conduct numerical experiments to validate our theoretical findings.
The software-defined network implements the separation of software and hardware, as well as a programmable network. however, the lack of consistent records of network data poses difficulties for network management, and heterogeneous device heterogeneity poses a hindrance to software-defined network interoperability. This paper summarizes the development status and existing problems of software-defined network, proposes a softwaredefined network data chain based on blockchain, realizes distributed consistent record of software-defined network data, and breaks the multi-vendor device isolation for fault recovery. Reduce the cost of network failure recovery and achieve unified scheduling of business capabilities.
In this paper, we introduce a Homogeneous Second-Order Descent Method (HSODM) using the homogenized quadratic approximation to the original function. The merit of homogenization is that only the leftmost eigenvector of a gradient-Hessian integrated matrix is computed at each iteration. Therefore, the algorithm is a single-loop method that does not need to switch to other sophisticated algorithms, and is easy to be implemented. We show that HSODM has a global convergence rate of $O(\epsilon^{-3/2})$ to find an $\epsilon$-approximate second-order stationary point, and has a local quadratic convergence rate under the standard assumptions. The numerical results demonstrate the advantage of the proposed method over other second-order methods.
Deploying reliable deep learning techniques in interdisciplinary applications needs learned models to output accurate and (even more importantly) explainable predictions. Existing approaches typically explicate network outputs in a post-hoc fashion, under an implicit assumption that faithful explanations come from accurate predictions/classifications. We have an opposite claim that explanations boost (or even determine) classification. That is, end-to-end learning of explanation factors to augment discriminative representation extraction could be a more intuitive strategy to inversely assure fine-grained explainability, e.g., in those neuroimaging and neuroscience studies with high-dimensional data containing noisy, redundant, and task-irrelevant information. In this paper, we propose such an explainable geometric deep network dubbed as NeuroExplainer, with applications to uncover altered infant cortical development patterns associated with preterm birth. Given fundamental cortical attributes as network input, our NeuroExplainer adopts a hierarchical attention-decoding framework to learn fine-grained attentions and respective discriminative representations to accurately recognize preterm infants from term-born infants at term-equivalent age. NeuroExplainer learns the hierarchical attention-decoding modules under subject-level weak supervision coupled with targeted regularizers deduced from domain knowledge regarding brain development. These prior-guided constraints implicitly maximizes the explainability metrics (i.e., fidelity, sparsity, and stability) in network training, driving the learned network to output detailed explanations and accurate classifications. Experimental results on the public dHCP benchmark suggest that NeuroExplainer led to quantitatively reliable explanation results that are qualitatively consistent with representative neuroimaging studies.
Nucleosides have been found to suffer in-source fragmentation (ISF) in electrospray ionization mass spectrometry, which leads to reduced sensitivity and ambiguous identification. In this work, a combination of theoretical calculations and nuclear magnetic resonance analysis revealed the key role of protonation at N3 near the glycosidic bond during ISF. Therefore, an ultrasensitive liquid chromatography-tandem mass spectrometry system for 5-formylcytosine detection was developed with 300 fold signal enhancement. Also, we established a MS1-only platform for nucleoside profiling and successfully identified sixteen nucleosides in the total RNA of MCF-7 cells. Taking ISF into account, we can realize analysis with higher sensitivity and less ambiguity, not only for nucleosides, but for other molecules with similar protonation and fragmentation behaviors.
Salmonella enterica remains one of the leading causes of foodborne bacterial disease. Retail meat is a major source of human salmonellosis. However, comparative genomic analyses of S. enterica isolates from retail meat from different sources in China are lacking. A total of 341 S. enterica strains were isolated from retail meat in sixteen districts of Beijing, China, at three different time points (January 1st, May 1st, and October 1st) in 2017. Comparative genomics was performed to investigate the genetic diversity, virulence and antimicrobial resistance gene (ARG) profiles of these isolates. The most common serotype was S. Enteritidis (203/341, 59.5%), which dominated among isolates from three different time points during the year. Laboratory retesting confirmed the accuracy of the serotyping results predicted by the Salmonella In Silico Typing Resource (SISTR) (96.5%). The pangenome of the 341 S. enterica isolates contained 13,931 genes, and the core genome contained 3,635 genes. Higher Salmonella phage 118970 sal3 (219/341, 64.2%) and Gifsy-2 (206/341, 60.4%) prevalence contributed to the diversity of the accessory genes, especially those with unknown functions. IncFII(S), IncX1, and IncFIB(S) plasmid replicons were more common in these isolates and were major sources of horizontally acquired foreign genes. The virulence gene profile showed fewer virulence genes associated with type III secretion systems in certain isolates from chicken. A total of 88 different ARGs were found in the 341 isolates. Three beta-lactamases, namely, bla CTX – M – 55 ( n = 15), bla CTX – M – 14 ( n = 11), and bla CTX – M – 65 ( n = 11), were more prevalent in retail meats. The emergence of qnrE1 and bla CTX – M – 123 indicated a potential increase in the prevalence of retail meats. After the prohibition of colistin in China, three and four isolates were positive for the colistin resistance genes mcr-1.1 and mcr-9 , respectively. Thus, we explored the evolution and genomic features of S. enterica isolates from retail meats in Beijing, China. The diverse ARGs of these isolates compromise food security and are a clinical threat.
Meat adulteration is currently a common practice worldwide. In China, adulteration of donkey meat products with the similar species (horse and mule/hinny) meat and mislabeling are becoming widespread concerns. In this study, a sensitive and species-specific duplex real-time PCR assay based on the simultaneous amplification of fragments of the creatine kinase muscle gene family, was developed and optimized for the identification of horse, donkey and mule /hinny species in raw and heat-processed meat products. Duplex real-time PCR results showed different fluorescence amplification curves for horse and donkey. Both kinds of fluorescence amplification curves appeared simultaneously for mule/hinny. The limit of detection (LOD) of the method was up to 0.01 ng /μL. The method and strategy developed in this study could be applied to detect the presence of adulterants from horse and mule /hinny meat in raw donkey meat and meat products.
The ADMM-based interior point (ABIP, Lin et al. 2021) method is a hybrid algorithm that effectively combines interior point method (IPM) and first-order methods to achieve a performance boost in large-scale linear optimization. Different from traditional IPM that relies on computationally intensive Newton steps, the ABIP method applies the alternating direction method of multipliers (ADMM) to approximately solve the barrier penalized problem. However, similar to other first-order methods, this technique remains sensitive to condition number and inverse precision. In this paper, we provide an enhanced ABIP method with multiple improvements. Firstly, we develop an ABIP method to solve the general linear conic optimization and establish the associated iteration complexity. Secondly, inspired by some existing methods, we develop different implementation strategies for ABIP method, which substantially improve its performance in linear optimization. Finally, we conduct extensive numerical experiments in both synthetic and real-world datasets to demonstrate the empirical advantage of our developments. In particular, the enhanced ABIP method achieves a 5.8x reduction in the geometric mean of run time on $105$ selected LP instances from Netlib, and it exhibits advantages in certain structured problems such as SVM and PageRank. However, the enhanced ABIP method still falls behind commercial solvers in many benchmarks, especially when high accuracy is desired. We posit that it can serve as a complementary tool alongside well-established solvers.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)