Improving the Seasonal Forecast of Summer Precipitation in Southeastern China Using a CycleGAN-based Deep Learning Bias Correction Method

Accurate seasonal precipitation forecasts,especially for extreme events,are crucial to preventing meteorological hazards and their potential impacts on national development,social activity,and security.However,the intensity of summer precipitation is often largely underestimated in many current dynamic models.This study uses a deep learning method called Cycle-Consistent Generative Adversarial Networks(CycleGAN)to improve the seasonal forecasts for June-JulyAugust precipitation in southeastern China by the Nanjing University of Information Science and Technology Climate Forecast System(NUIST-CFS 1.0).The results suggest that the CycleGAN-based model significantly improves the accuracy in predicting the spatiotemporal distribution of summer precipitation compared to the traditional quantile mapping(QM)method.Using the unpaired bias-correction model,we can also obtain advanced forecasts of the frequency,intensity,and duration of extreme precipitation events over the dynamic model predictions.This study expands the potential applications of deep learning models toward improving seasonal precipitation forecasts.

ResoNet:Robust and Explainable ENSO Forecasts with Hybrid Convolution and Transformer Networks

Recent studies have shown that deep learning(DL)models can skillfully forecast El Niño–Southern Oscillation(ENSO)events more than 1.5 years in advance.However,concerns regarding the reliability of predictions made by DL methods persist,including potential overfitting issues and lack of interpretability.Here,we propose ResoNet,a DL model that combines CNN(convolutional neural network)and transformer architectures.This hybrid architecture enables our model to adequately capture local sea surface temperature anomalies as well as long-range inter-basin interactions across oceans.We show that ResoNet can robustly predict ENSO at lead times of 19 months,thus outperforming existing approaches in terms of the forecast horizon.According to an explainability method applied to ResoNet predictions of El Niño and La Niña from 1-to 18-month leads,we find that it predicts the Niño-3.4 index based on multiple physically reasonable mechanisms,such as the recharge oscillator concept,seasonal footprint mechanism,and Indian Ocean capacitor effect.Moreover,we demonstrate for the first time that the asymmetry between El Niño and La Niña development can be captured by ResoNet.Our results could help to alleviate skepticism about applying DL models for ENSO prediction and encourage more attempts to discover and predict climate phenomena using AI methods.

LÉVY AREA ANALYSIS AND PARAMETER ESTIMATION FOR FOU PROCESSES VIA NON-GEOMETRIC ROUGH PATH THEORY

This paper addresses the estimation problem of an unknown drift parameter matrix for a fractional Ornstein-Uhlenbeck process in a multi-dimensional setting.To tackle this problem,we propose a novel approach based on rough path theory that allows us to construct pathwise rough path estimators from both continuous and discrete observations of a single path.Our approach is particularly suitable for high-frequency data.To formulate the parameter estimators,we introduce a theory of pathwise Itôintegrals with respect to fractional Brownian motion.By establishing the regularity of fractional Ornstein-Uhlenbeck processes and analyzing the long-term behavior of the associated Lévy area processes,we demonstrate that our estimators are strongly consistent and pathwise stable.Our findings offer a new perspective on estimating the drift parameter matrix for fractional Ornstein-Uhlenbeck processes in multi-dimensional settings,and may have practical implications for fields including finance,economics,and engineering.

Investigation of Power Density Amplification in Stacked Triboelectric Nanogenerators

In engineering practice,the output performance of contact separation TENGs(CS-TENGs)increases with the increase of tribo-pair area,which includes increasing the size of single layer CS-TENGs(SCS-TENGs)or the number of units(zigzag TENGs).However,such two strategies show significant differences in output power and power density.In this study,to seek a universal CS-TENG design solution,the output performance of a SCS-TENG and a zigzag TENG(Z-TENG)is systematically compared,including voltage,current,transferred charge,instantaneous power density,and charging power density.The relationship between contact area and output voltages is explored,and the output voltage equation is fitted.The experimental results reveal that SCS-TENGs yield better performance than Z-TENGs in terms of voltage,power,and power density under the same total contact area.Z-TENGs show energy loss during the transfer of mechanical energy,and such loss is aggravated by the increasing number of units.The instantaneous peak power of the SCS-TENG is up to 22 times that of the Z-TENG(45 cm^(2)).Furthermore,the power density of capacitor charging of SCS-TENGs is 131%of that of Z-TENGs,which are relatively close.Z-TENG is a feasible alternative when the working space is limited.

Deep learning-based automatic pipeline system for predicting lateral cervical lymph node metastasis in patients with papillary thyroid carcinoma using computed tomography:A multi-center study

Objective:The assessment of lateral lymph node metastasis(LLNM)in patients with papillary thyroid carcinoma(PTC)holds great significance.This study aims to develop and evaluate a deep learning-based automatic pipeline system(DLAPS)for diagnosing LLNM in PTC using computed tomography(CT).Methods:A total of 1,266 lateral lymph nodes(LLNs)from 519 PTC patients who underwent CT examinations from January 2019 to November 2022 were included and divided into training and validation set,internal test set,pooled external test set,and prospective test set.The DLAPS consists of an auto-segmentation network based on RefineNet model and a classification network based on ensemble model(ResNet,Xception,and DenseNet).The performance of the DLAPS was compared with that of manually segmented DL models,the clinical model,and Node Reporting and Data System(Node-RADS).The improvement of radiologists’diagnostic performance under the DLAPS-assisted strategy was explored.In addition,bulk RNA-sequencing was conducted based on 12 LLNs to reveal the underlying biological basis of the DLAPS.Results:The DLAPS yielded good performance with area under the receiver operating characteristic curve(AUC)of 0.872,0.910,and 0.822 in the internal,pooled external,and prospective test sets,respectively.The DLAPS significantly outperformed clinical models(AUC 0.731,P<0.001)and Node-RADS(AUC 0.602,P<0.001)in the internal test set.Moreover,the performance of the DLAPS was comparable to that of the manually segmented deep learning(DL)model with AUCs ranging 0.814−0.901 in three test sets.Furthermore,the DLAPSassisted strategy improved the performance of radiologists and enhanced inter-observer consistency.In clinical situations,the rate of unnecessary LLN dissection decreased from 33.33%to 7.32%.Furthermore,the DLAPS was associated with the cell-cell conjunction in the microenvironment.Conclusions:Using CT images from PTC patients,the DLAPS could effectively segment and classify LLNs non-invasively,and this system had a good generalization ability and clinical applicability.

面向位置大数据的快速密度聚类算法

面向位置大数据聚类,提出了一种简单但高效的快速密度聚类算法CBSCAN,以快速发现位置大数据中任意形状的聚类簇模式和噪声.首先,定义了Cell网格概念,并提出了基于Cell的距离分析理论,利用该距离分析,无需距离计算,可快速确定高密度区域的核心点和密度相连关系;其次,给出了网格簇定义,将基于位置点的密度簇映射成基于网格的密度簇,利用排他网格与相邻网格的密度关系,可快速确定网格簇的包含网格;第三,利用基于Cell的距离分析理论和网格簇概念,实现了一个快速密度聚类算法,将DBSCAN基于数据点的密度扩展聚类转换成基于Cell的密度扩展聚类,极大地减少高密度区域的距离计算,利用位置数据的内在特性提高了聚类效率;最后,在基准测试数据上验证了所提算法的聚类效果,在位置大数据上的实验结果统计显示,与DBSCAN、PR-Tree索引和Grid索引优化的DBSCAN相比,CBSCAN分别平均提升了525倍、30倍和11倍效率.

Motor Imagery and Error Related Potential Induced Position Control of a Robotic Arm

The paper introduces an electroencephalography(EEG) driven online position control scheme for a robot arm by utilizing motor imagery to activate and error related potential(ErrP) to stop the movement of the individual links, following a fixed(pre-defined) order of link selection. The right(left)hand motor imagery is used to turn a link clockwise(counterclockwise) and foot imagery is used to move a link forward. The occurrence of ErrP here indicates that the link under motion crosses the visually fixed target position, which usually is a plane/line/point depending on the desired transition of the link across 3D planes/around 2D lines/along 2D lines respectively. The imagined task about individual link's movement is decoded by a classifier into three possible class labels: clockwise, counterclockwise and no movement in case of rotational movements and forward, backward and no movement in case of translational movements. One additional classifier is required to detect the occurrence of the ErrP signal, elicited due to visually inspired positional link error with reference to a geometrically selected target position. Wavelet coefficients and adaptive autoregressive parameters are extracted as features for motor imagery and ErrP signals respectively. Support vector machine classifiers are used to decode motor imagination and ErrP with high classification accuracy above 80%. The average time taken by the proposed scheme to decode and execute control intentions for the complete movement of three links of a robot is approximately33 seconds. The steady-state error and peak overshoot of the proposed controller are experimentally obtained as 1.1% and4.6% respectively.

Artificial intelligence for assisting cancer diagnosis and treatment in the era of precision medicine

Over the past decade,artificial intelligence(AI)has contributed substantially to the resolution of various medical problems,including cancer.Deep learning(DL),a subfield of AI,is characterized by its ability to perform automated feature extraction and has great power in the assimilation and evaluation of large amounts of complicated data.On the basis of a large quantity of medical data and novel computational technologies,AI,especially DL,has been applied in various aspects of oncology research and has the potential to enhance cancer diagnosis and treatment.These applications range from early cancer detection,diagnosis,classification and grading,molecular characterization of tumors,prediction of patient outcomes and treatment responses,personalized treatment,automatic radiotherapy workflows,novel anti-cancer drug discovery,and clinical trials.In this review,we introduced the general principle of AI,summarized major areas of its application for cancer diagnosis and treatment,and discussed its future directions and remaining challenges.As the adoption of AI in clinical use is increasing,we anticipate the arrival of AI-powered cancer care.

Scribble-Supervised Video Object Segmentation

Recently,video object segmentation has received great attention in the computer vision community.Most of the existing methods heavily rely on the pixel-wise human annotations,which are expensive and time-consuming to obtain.To tackle this problem,we make an early attempt to achieve video object segmentation with scribble-level supervision,which can alleviate large amounts of human labor for collecting the manual annotation.However,using conventional network architectures and learning objective functions under this scenario cannot work well as the supervision information is highly sparse and incomplete.To address this issue,this paper introduces two novel elements to learn the video object segmentation model.The first one is the scribble attention module,which captures more accurate context information and learns an effective attention map to enhance the contrast between foreground and background.The other one is the scribble-supervised loss,which can optimize the unlabeled pixels and dynamically correct inaccurate segmented areas during the training stage.To evaluate the proposed method,we implement experiments on two video object segmentation benchmark datasets,You Tube-video object segmentation(VOS),and densely annotated video segmentation(DAVIS)-2017.We first generate the scribble annotations from the original per-pixel annotations.Then,we train our model and compare its test performance with the baseline models and other existing works.Extensive experiments demonstrate that the proposed method can work effectively and approach to the methods requiring the dense per-pixel annotations.

人工智能在化学逆合成中的应用

近年来,人们对通过人工智能(AI)技术解决化学逆合成预测问题产生了巨大的兴趣。与化学家和基于规则的专家系统进行的逆合成预测不同,AI驱动的逆合成预测自动从现成的实验数据集中学习化学知识,以预测反应和逆合成路径。这提供了解决许多传统挑战的机会,包括专业知识过于繁杂、路线的长度过大和计算成本高。本文描述了当前AI驱动的逆合成预测的概况。我们首先讨论了化学逆合成的数学定义,并回顾了这个问题中的研究挑战。然后,我们回顾相关的AI技术和最新进度,以实现逆合成预测。此外,我们提供了逆合成系统不同部分的全面分类,并调查了AI方法如何重塑每个子模块。最后,我们讨论了未来有希望的研究领域。

Benefits Analysis of Beam Hopping in Satellite Mobile System with Unevenly Distributed Traffic

Satellite mobile system and space-airground integrated network have a prominent superiority in global coverage which plays a critical role in remote and non-land regions, as well as emergency communications. However, due to the gradual angle attenuations of the satellite antennas, it is difficult to achieve full frequency multiplex among different beams as terrestrial 5G network. Multi-color frequency reuse is widely adopted in both academic and industry. Beam hopping scheme has attracted the attention of researchers recently due to the allocation flexibility. In this paper, we focus on analyzing the performance benefits of beam hopping compared with multi-color frequency reuse scheme in non-uniform user and traffic distributions in satellite system. Aerial networks are also introduced to form a space-airground integrated network for coverage enhancement,and the capacity improvement is analyzed. Besides,additional improved techniques are provided to make comprehensive analysis and comparisons. Theoretical analysis and simulation results indicate that the beam hopping scheme has a prominent superiority in the system capacity compared with the traditional multicolor frequency reuse scheme in both satellite mobile system and future space-air-ground integrated network.

Detection and classification of breast lesions using multiple information on contrast-enhanced mammography by a multiprocess deep-learning system: A multicenter study

Objective: Accurate detection and classification of breast lesions in early stage is crucial to timely formulate effective treatments for patients. We aim to develop a fully automatic system to detect and classify breast lesions using multiple contrast-enhanced mammography(CEM) images.Methods: In this study, a total of 1,903 females who underwent CEM examination from three hospitals were enrolled as the training set, internal testing set, pooled external testing set and prospective testing set. Here we developed a CEM-based multiprocess detection and classification system(MDCS) to perform the task of detection and classification of breast lesions. In this system, we introduced an innovative auxiliary feature fusion(AFF)algorithm that could intelligently incorporates multiple types of information from CEM images. The average freeresponse receiver operating characteristic score(AFROC-Score) was presented to validate system’s detection performance, and the performance of classification was evaluated by area under the receiver operating characteristic curve(AUC). Furthermore, we assessed the diagnostic value of MDCS through visual analysis of disputed cases,comparing its performance and efficiency with that of radiologists and exploring whether it could augment radiologists’ performance.Results: On the pooled external and prospective testing sets, MDCS always maintained a high standalone performance, with AFROC-Scores of 0.953 and 0.963 for detection task, and AUCs for classification were 0.909[95% confidence interval(95% CI): 0.822-0.996] and 0.912(95% CI: 0.840-0.985), respectively. It also achieved higher sensitivity than all senior radiologists and higher specificity than all junior radiologists on pooled external and prospective testing sets. Moreover, MDCS performed superior diagnostic efficiency with an average reading time of 5 seconds, compared to the radiologists’ average reading time of 3.2 min. The average performance of all radiologists was also improved to varying degrees with MDCS assistance.Conclusions: MDCS demonstrated excellent performance in the detection and classification of breast lesions,and greatly enhanced the overall performance of radiologists.

Traveling Salesman Problem Using an Enhanced Hybrid Swarm Optimization Algorithm

The traveling salesman problem( TSP) is a well-known combinatorial optimization problem as well as an NP-complete problem. A dynamic multi-swarm particle swarm optimization and ant colony optimization( DMPSO-ACO) was presented for TSP.The DMPSO-ACO combined the exploration capabilities of the dynamic multi-swarm particle swarm optimizer( DMPSO) and the stochastic exploitation of the ant colony optimization( ACO) for solving the traveling salesman problem. In the proposed hybrid algorithm,firstly,the dynamic swarms,rapidity of the PSO was used to obtain a series of sub-optimal solutions through certain iterative times for adjusting the initial allocation of pheromone in ACO. Secondly,the positive feedback and high accuracy of the ACO were employed to solving whole problem. Finally,to verify the effectiveness and efficiency of the proposed hybrid algorithm,various scale benchmark problems were tested to demonstrate the potential of the proposed DMPSO-ACO algorithm. The results show that DMPSO-ACO is better in the search precision,convergence property and has strong ability to escape from the local sub-optima when compared with several other peer algorithms.

Using Event-Based Method to Estimate Cybersecurity Equilibrium

Estimating the global state of a networked system is an important problem in many application domains.The classical approach to tackling this problem is the periodic(observation)method,which is inefficient because it often observes states at a very high frequency.This inefficiency has motivated the idea of event-based method,which leverages the evolution dynamics in question and makes observations only when some rules are triggered(i.e.,only when certain conditions hold).This paper initiates the investigation of using the event-based method to estimate the equilibrium in the new application domain of cybersecurity,where equilibrium is an important metric that has no closed-form solutions.More specifically,the paper presents an event-based method for estimating cybersecurity equilibrium in the preventive and reactive cyber defense dynamics,which has been proven globally convergent.The presented study proves that the estimated equilibrium from our trigger rule i)indeed converges to the equilibrium of the dynamics and ii)is Zeno-free,which assures the usefulness of the event-based method.Numerical examples show that the event-based method can reduce 98%of the observation cost incurred by the periodic method.In order to use the event-based method in practice,this paper investigates how to bridge the gap between i)the continuous state in the dynamics model,which is dubbed probability-state because it measures the probability that a node is in the secure or compromised state,and ii)the discrete state that is often encountered in practice,dubbed sample-state because it is sampled from some nodes.This bridge may be of independent value because probability-state models have been widely used to approximate exponentially-many discrete state systems.

Construction of Influenza Early Warning Model Based on Combinatorial Judgment Classifier:A Case Study of Seasonal Influenza in Hong Kong

Objective:The annual influenza epidemic is a heavy burden on the health care system,and has increasingly become a major public health problem in some areas,such as Hong Kong(China).Therefore,based on a variety of machine learning methods,and considering the seasonal influenza in Hong Kong,the study aims to establish a Combinatorial Judgment Classifier(CJC)model to classify the epidemic trend and improve the accuracy of influenza epidemic early warning.

Accurate querying of frequent subgraphs in power grid graph data

With the development of information technology, the amount of power grid topology data has gradually increased. Therefore, accurate querying of this data has become particularly important. Several researchers have chosen different indexing methods in the filtering stage to obtain more optimized query results because currently there is no uniform and efficient indexing mechanism that achieves good query results. In the traditional algorithm, the hash table for index storage is prone to "collision" problems, which decrease the index construction efficiency. Aiming at the problem of quick index entry, based on the construction of frequent subgraph indexes, a method of serialized storage optimization based on multiple hash tables is proposed. This method mainly uses the exploration sequence to make the keywords evenly distributed; it avoids conflicts of the stored procedure and performs a quick search of the index. The proposed algorithm mainly adopts the "filterverify" mechanism; in the filtering stage, the index is first established offline, and then the frequent subgraphs are found using the "contains logic" rule to obtain the candidate set. Experimental results show that this method can reduce the time and scale of candidate set generation and improve query efficiency.

Establishing and validating a spotted tongue recognition and extraction model based on multiscale convolutional neural network

Objective In tongue diagnosis,the location,color,and distribution of spots can be used to speculate on the viscera and severity of the heat evil.This work focuses on the image analysis method of artificial intelligence(AI)to study the spotted tongue recognition of traditional Chinese medicine(TCM).Methods A model of spotted tongue recognition and extraction is designed,which is based on the principle of image deep learning and instance segmentation.This model includes multiscale feature map generation,region proposal searching,and target region recognition.Firstly,deep convolution network is used to build multiscale low-and high-abstraction feature maps after which,target candidate box generation algorithm and selection strategy are used to select high-quality target candidate regions.Finally,classification network is used for classifying target regions and calculating target region pixels.As a result,the region segmentation of spotted tongue is obtained.Under non-standard illumination conditions,various tongue images were taken by mobile phones,and experiments were conducted.Results The spotted tongue recognition achieved an area under curve(AUC)of 92.40%,an accuracy of 84.30%with a sensitivity of 88.20%,a specificity of 94.19%,a recall of 88.20%,a regional pixel accuracy index pixel accuracy(PA)of 73.00%,a mean pixel accuracy(m PA)of73.00%,an intersection over union(Io U)of 60.00%,and a mean intersection over union(mIo U)of 56.00%.Conclusion The results of the study verify that the model is suitable for the application of the TCM tongue diagnosis system.Spotted tongue recognition via multiscale convolutional neural network(CNN)would help to improve spot classification and the accurate extraction of pixels of spot area as well as provide a practical method for intelligent tongue diagnosis of TCM.

Comparison of Production-Triggering Strategies for Manufacturing Enterprises under Random Orders

Although new technologies have been deeply applied in manufacturing systems,manufacturing enterprises are still encountering difficulties in maintaining efficient and flexible production due to the random arrivals of diverse customer requirements.Fast order delivery and low inventory cost are fundamentally contradictory to each other.How to make a suitable production-triggering strategy is a critical issue for an enterprise to maintain a high level of competitiveness in a dynamic environment.In this paper,we focus on production-triggering strategies for manufacturing enterprises to satisfy randomly arriving orders and reduce inventory costs.Unified theoretical models and simulation models of different production strategies are proposed,including time-triggered strategies,event-triggered strategies,and hybrid-triggered strategies.In each model,both part-production-triggering strategies and product-assembly-triggering strategies are considered and implemented.The time-triggered models and hybrid-triggered models also consider the impact of the period on system performance.The results show that hybrid-triggered and time-triggered strategies yield faster order delivery and lower inventory costs than event-triggered strategies if the period is set appropriately.

Performance of statistical and machine learning-based methods for predicting biogeographical patterns of fungal productivity in forest ecosystems

Background:The prediction of biogeographical patterns from a large number of driving factors with complex interactions,correlations and non-linear dependences require advanced analytical methods and modeling tools.This study compares different statistical and machine learning-based models for predicting fungal productivity biogeographical patterns as a case study for the thorough assessment of the performance of alternative modeling approaches to provide accurate and ecologically-consistent predictions.Methods:We evaluated and compared the performance of two statistical modeling techniques,namely,generalized linear mixed models and geographically weighted regression,and four techniques based on different machine learning algorithms,namely,random forest,extreme gradient boosting,support vector machine and artificial neural network to predict fungal productivity.Model evaluation was conducted using a systematic methodology combining random,spatial and environmental blocking together with the assessment of the ecological consistency of spatially-explicit model predictions according to scientific knowledge.Results:Fungal productivity predictions were sensitive to the modeling approach and the number of predictors used.Moreover,the importance assigned to different predictors varied between machine learning modeling approaches.Decision tree-based models increased prediction accuracy by more than 10%compared to other machine learning approaches,and by more than 20%compared to statistical models,and resulted in higher ecological consistence of the predicted biogeographical patterns of fungal productivity.Conclusions:Decision tree-based models were the best approach for prediction both in sampling-like environments as well as in extrapolation beyond the spatial and climatic range of the modeling data.In this study,we show that proper variable selection is crucial to create robust models for extrapolation in biophysically differentiated areas.This allows for reducing the dimensions of the ecosystem space described by the predictors of the models,resulting in higher similarity between the modeling data and the environmental conditions over the whole study area.When dealing with spatial-temporal data in the analysis of biogeographical patterns,environmental blocking is postulated as a highly informative technique to be used in cross-validation to assess the prediction error over larger scales.