Gradient Descent Ascent for Minimax Problems on Riemannian Manifolds.

In the paper, we study a class of useful minimax problems on Riemanian manifolds and propose a class of effective Riemanian gradient-based methods to solve these minimax problems. Specifically, we propose an effective Riemannian gradient descent ascent (RGDA) algorithm for the deterministic minimax optimization. Moreover, we prove that our RGDA has a sample complexity of O(κ2ϵ-2) for finding an ϵ-stationary solution of the Geodesically-Nonconvex Strongly-Concave (GNSC) minimax problems, where κ denotes the condition number. At the same time, we present an effective Riemannian stochastic gradient descent ascent (RSGDA) algorithm for the stochastic minimax optimization, which has a sample complexity of O(κ4ϵ-4) for finding an ϵ-stationary solution. To further reduce the sample complexity, we propose an accelerated Riemannian stochastic gradient descent ascent (Acc-RSGDA) algorithm based on the momentum-based variance-reduced technique. We prove that our Acc-RSGDA algorithm achieves a lower sample complexity of ~O(κ4ϵ-3) in searching for an ϵ-stationary solution of the GNSC minimax problems. Extensive experimental results on the robust distributional optimization and robust Deep Neural Networks (DNNs) training over Stiefel manifold demonstrate efficiency of our algorithms.

State Causality and Adaptive Covariance Decomposition Based Time Series Forecasting.

Time series forecasting is a very vital research topic. The scale of time series in numerous industries has risen considerably in recent years as a result of the advancement of information technology. However, the existing algorithms pay little attention to generating large-scale time series. This article designs a state causality and adaptive covariance decomposition-based time series forecasting method (SCACD). As an observation sequence, the majority of time series is generated under the influence of hidden states. First, SCACD builds neural networks to adaptively estimate the mean and covariance matrix of latent variables; Then, SCACD employs causal convolution to forecast the distribution of future latent variables; Lastly, to avoid loss of information, SCACD applies a sampling approach based on Cholesky decomposition to generate latent variables and observation sequences. Compared to existing outstanding time series prediction models on six real datasets, the model can achieve long-term forecasting while also being lighter, and the forecasting accuracy is improved in the great majority of the prediction tasks.

Investigation and Analysis of Infection among Inpatients in a Tertiary Hospital in Shanghai.

Objectives: This study is aimed at obtaining information about the prevalence of nosocomial infections (NIs) and the use of antibiotics in hospitalized patients and providing relevant references for further understanding, preventing, and controlling NIs. Methods: The medical records of adult patients admitted to a hospital in Shanghai from November to December 2021 were analyzed. The patients were divided into the NI group, community-acquired infection (CAI) group, and uninfected or healed group according to their infection status. The survey results were summarized and analyzed. Results: A total of 1485 patients were investigated, including 115 patients in the NI group, 172 patients in the CAI group, and 1198 patients in the uninfected or healed group. In the NI group, the main infection site was intra-abdominal tissue (49.6%), followed by lower respiratory tract (unrelated to application of catheters) (13%). There were 73 pathogens detected in the samples submitted from the NI group, mainly including 8 cases (11%) of Escherichia coli, 9 cases (12%) of Klebsiella pneumoniae, and 40 cases (55%) of negative microbiological test results. Thirteen of 115 patients with NIs had infections with drug-resistant bacteria, including 9 cases (69.2%) of CRE (carbapenem-resistant Enterobacteriaceae), 2 cases (15.38%) of VRE (vancomycin-resistant Enterococcus), 1 case (7.69%) of MRSA (methicillin-resistant Staphylococcus aureus), and 1 case (7.69%) of CRAB (carbapenem-resistant Acinetobacter baumannii). In terms of medication, single drug use accounted for the majority of the NI group (66.3%) and CAI group (60.6%); both groups had less frequent quadruple drugs. In the uninfected or healed group, single drug occupied 92.0%, and dual drug use accounted for 8.0%. Cefoperazone/sulbactam was the most commonly used antibacterial drug in the NI group (18.0%) and CAI group (17.6%), and piperacillin/tazobactam accounted for 14.0% and 17.6% in the two groups, respectively. In the uninfected or healed group, cefuroxime accounted for 59.8%, followed by cefoperazone/sulbactam (13.3%). Conclusion: This study provides a scientific basis for effective control of NIs. Strict implementation of aseptic techniques, reduction of invasive operations, and rational use of anti-infective drugs can minimize the incidence of nosocomial infection to ultimately achieve effective prevention and control of NIs.

Value of 5-Hydroxymethylcytosine in HBV-Carrying High-Risk Hepatocellular Carcinoma Population: An Evaluation Based on Differential Analysis.

Objective: To clarify the application value of 5-hydroxymethylcytosine (5hmC) in evaluating the progression of chronic hepatitis B (CHB) to hepatocellular carcinoma (HCC) based on difference analysis. Methods: A total of 180 patients were enrolled. Among them, 84 patients with chronic hepatitis B virus (HBV) infection while no progression to hepatocellular carcinoma (HCC) were included in the control group (CG), and 96 patients with HCC developed from HBV infection were included in the research group (RG). Two-thirds of the samples were used in the training set and 1/3 samples in the validation set to detect the level of 5hmC in both groups based on the modified nano-hmC-Seal technique. The expression levels of 5hmC-related genes TET2 and TET3 were quantified by qPCR, and the correlation between TET3 and 5hmC was analyzed by Pearson's correlation coefficients. Receiver operating characteristic (ROC) curves were drawn to evaluate the application value of the TET3-based 5hmC prediction model in the early diagnosis of HCC. Results: (i) The expression of 5hmC in RG was lower than that in CG, no matter in the training set or the validation set. (ii) 5hmC was significantly enriched in the region between the transcription initiation site and the transcription end site but was depleted in the flanking region. (iii) 5hmC-related genes TET2 and TET3 were significantly downregulated in HCC patients, whether in the training set or the validation set. (iv) In both the training and validation sets, TET3 showed a positive association with 5hmC. (v) ROC analysis results showed that the 5hmC prediction model could be used to predict the progression of CHB to HCC (training set: AUC = 0.81, 0.729-0.893; validation set: AUC = 0.84, 0.739-0.936). Conclusions: TET3 expression based on 5hmC sequencing is a landmark molecule for evaluating the progression of HCC in CHB patients, which is worthy of further study and promotion.

Riemannian gradient methods for stochastic composition problems.

In the paper, we study a class of novel stochastic composition optimization problems over Riemannian manifold, which have been raised by multiple emerging machine learning applications such as distributionally robust learning in Riemannian manifold setting. To solve these composition problems, we propose an effective Riemannian compositional gradient (RCG) algorithm, which has a sample complexity of O(ϵ-4) for finding an ϵ-stationary point. To further reduce sample complexity, we propose an accelerated momentum-based Riemannian compositional gradient (M-RCG) algorithm. Moreover, we prove that the M-RCG obtains a lower sample complexity of Õ(ϵ-3) without large batches, which achieves the best known sample complexity for its Euclidean counterparts. Extensive numerical experiments on training deep neural networks (DNNs) over Stiefel manifold and learning principal component analysis (PCA) over Grassmann manifold demonstrate effectiveness of our proposed algorithms. To the best of our knowledge, this is the first study of the composition optimization problems over Riemannian manifold.

Faster Stochastic Quasi-Newton Methods.

Stochastic optimization methods have become a class of popular optimization tools in machine learning. Especially, stochastic gradient descent (SGD) has been widely used for machine learning problems, such as training neural networks, due to low per-iteration computational complexity. In fact, the Newton or quasi-newton (QN) methods leveraging the second-order information are able to achieve a better solution than the first-order methods. Thus, stochastic QN (SQN) methods have been developed to achieve a better solution efficiently than the stochastic first-order methods by utilizing approximate second-order information. However, the existing SQN methods still do not reach the best known stochastic first-order oracle (SFO) complexity. To fill this gap, we propose a novel faster stochastic QN method (SpiderSQN) based on the variance reduced technique of SIPDER. We prove that our SpiderSQN method reaches the best known SFO complexity of O(n+n1/2ϵ-2) in the finite-sum setting to obtain an ϵ -first-order stationary point. To further improve its practical performance, we incorporate SpiderSQN with different momentum schemes. Moreover, the proposed algorithms are generalized to the online setting, and the corresponding SFO complexity of O(ϵ-3) is developed, which also matches the existing best result. Extensive experiments on benchmark data sets demonstrate that our new algorithms outperform state-of-the-art approaches for nonconvex optimization.

Hepatocellular carcinoma with worsened hypoglycemia after transarterial chemoembolization: a case report and systematic review.

Non-islet cell tumor hypoglycemia (NICTH) is an extremely uncommon and serious complication of hepatocellular carcinoma (HCC). Here, we reported a case of a 47-year-old male patient with moderate to poorly differentiated HCC complicated by hypoglycemia that worsened after transarterial chemoembolization (TACE). The patient was admitted into The First Affiliated Hospital of Naval Medical University due to fatigue, nausea, dizziness and passage of tea colored urine. He was diagnosed with NICTH induced by HCC according to CT scanning and laboratory tests. TACE was used as the primary therapy but the hypoglycemia worsened afterward. Then the patient received a liver transplantation as a possible radical cure and hypoglycemia was resolved. We systematically review the management of hypoglycemia caused by HCC and the results show that patients undergoing treatment that mainly alleviate tumor burdens obtained a significantly higher response rate than patients undergoing therapies mainly regulating biologic functions (50.0% vs 27.3%). Cytoreductive surgery, TACE and radiotherapy which aimed to alleviate tumor burdens are effective therapies have great potential, but the risk of hypoglycemic deterioration requires particular attention when using these treatments, especially with TACE.

Joint Estimation of Multiple Conditional Gaussian Graphical Models.

In this paper, we propose a joint conditional graphical Lasso to learn multiple conditional Gaussian graphical models, also known as Gaussian conditional random fields, with some similar structures. Our model builds on the maximum likelihood method with the convex sparse group Lasso penalty. Moreover, our model is able to model multiple multivariate linear regressions with unknown noise covariances via a convex formulation. In addition, we develop an efficient approximated Newton's method for optimizing our model. Theoretically, we establish the asymptotic properties of our model on consistency and sparsistency under the high-dimensional settings. Finally, extensive numerical results on simulations and real data sets demonstrate that our method outperforms the compared methods on structure recovery and structured output prediction. To the best of our knowledge, the joint learning of multiple multivariate regressions with unknown covariance is first studied.

Fluorescence sensing of ADP over ATP and PPi in 100% aqueous solution.

An anthracene-bridged dinuclear zinc(ii)-dipicolylamine complex was found to show high selectivity for ADP with a significant fluorescence enhancement over ATP, PPi and other common analytes in 100% aqueous solution. This complex can be used for fluorescence detection of ADP in living cells and for monitoring the activity of kinases.

Joint Learning of Multiple Sparse Matrix Gaussian Graphical Models.

We consider joint learning of multiple sparse matrix Gaussian graphical models and propose the joint matrix graphical Lasso to discover the conditional independence structures among rows (columns) in the matrix variable under distinct conditions. The proposed approach borrows strength across the different graphical models and is based on the maximum likelihood with penalized row and column precision matrices, respectively. In particular, our model is more parsimonious and flexible than the joint vector graphical models. Furthermore, we establish the asymptotic properties of our model on consistency and sparsistency. And the asymptotic analysis shows that our model enjoys a better convergence rate than the joint vector graphical models. Extensive simulation experiments demonstrate that our methods outperform state-of-the-art methods in identifying graphical structures and estimating precision matrices. Moreover, the effectiveness of our methods is also illustrated via a real data set analysis. Sparsistency is shorthand for consistency of the sparsity pattern of the parameters.

Near-infrared fluorescent probe for detection of thiophenols in water samples and living cells.

The development of probes for rapid, selective, and sensitive detection of the highly toxic thiophenols is of great importance in both environmental and biological science. Despite the appealing advantages of near-infrared (NIR) fluorescent detection, no NIR fluorescent probes have been reported for thiophenols to date. Using the chemical properties of thiophenols that are able to cleave sulfonamide selectively and efficiently under mild conditions, we herein report a dicyanomethylene-benzopyran (DCMB)-based NIR fluorescent probe for thiophenols. This probe features remarkable large Stokes shift and shows a rapid, highly selective, and sensitive detection process for thiophenols with significant NIR fluorescent turn-on responses. The potential applications of this new NIR fluorescent probe were demonstrated by the quantitative detection of thiophenol in real water samples and by fluorescent imaging of thiophenol in living cells.

Aldehyde group assisted thiolysis of dinitrophenyl ether: a new promising approach for efficient hydrogen sulfide probes.

We report herein a new approach, which combines fast nucleophilic addition of H2S to an aldehyde group and the subsequent intramolecular thiolysis of dinitrophenyl ether, and can be used to develop efficient and effective H2S probes.

Introducing ligand-based hydrogen bond donors to a receptor: both selectivity and binding affinity for anion recognition in water can be improved.

Introducing hydrogen bond donors to a receptor was found to be an effective approach to improve both its selectivity and binding affinity for pyrophosphate in water. The crystal structure of Zn3-ADP complex showed the improvements come from the combination of H-bonding and metal coordination in a manner similar to many metalloenzymes.