The sulfation and decomposition process has proven effective in selectively extracting lithium from lepidolite.It is essential to clarify the thermochemical behavior and kinetic parameters of decomposition reactions.A...The sulfation and decomposition process has proven effective in selectively extracting lithium from lepidolite.It is essential to clarify the thermochemical behavior and kinetic parameters of decomposition reactions.Accordingly,comprehensive kinetic study by employing thermalgravimetric analysis at various heating rates was presented in this paper.Two main weight loss regions were observed during heating.The initial region corresponded to the dehydration of crystal water,whereas the subsequent region with overlapping peaks involved complex decomposition reactions.The overlapping peaks were separated into two individual reaction peaks and the activation energy of each peak was calculated using isoconversional kinetics methods.The activation energy of peak 1 exhibited a continual increase as the reaction conversion progressed,while that of peak 2 steadily decreased.The optimal kinetic models,identified as belonging to the random nucleation and subsequent growth category,provided valuable insights into the mechanism of the decomposition reactions.Furthermore,the adjustment factor was introduced to reconstruct the kinetic mechanism models,and the reconstructed models described the kinetic mechanism model more accurately for the decomposition reactions.This study enhanced the understanding of the thermochemical behavior and kinetic parameters of the lepidolite sulfation product decomposition reactions,further providing theoretical basis for promoting the selective extraction of lithium.展开更多
Open caissons are widely used in foundation engineering because of their load-bearing efficiency and adaptability in diverse soil conditions.However,accurately predicting their undrained bearing capacity in layered so...Open caissons are widely used in foundation engineering because of their load-bearing efficiency and adaptability in diverse soil conditions.However,accurately predicting their undrained bearing capacity in layered soils remains a complex challenge.This study presents a novel application of five ensemble machine(ML)algorithms-random forest(RF),gradient boosting machine(GBM),extreme gradient boosting(XGBoost),adaptive boosting(AdaBoost),and categorical boosting(CatBoost)-to predict the undrained bearing capacity factor(Nc)of circular open caissons embedded in two-layered clay on the basis of results from finite element limit analysis(FELA).The input dataset consists of 1188 numerical simulations using the Tresca failure criterion,varying in geometrical and soil parameters.The FELA was performed via OptumG2 software with adaptive meshing techniques and verified against existing benchmark studies.The ML models were trained on 70% of the dataset and tested on the remaining 30%.Their performance was evaluated using six statistical metrics:coefficient of determination(R²),mean absolute error(MAE),root mean squared error(RMSE),index of scatter(IOS),RMSE-to-standard deviation ratio(RSR),and variance explained factor(VAF).The results indicate that all the models achieved high accuracy,with R²values exceeding 97.6%and RMSE values below 0.02.Among them,AdaBoost and CatBoost consistently outperformed the other methods across both the training and testing datasets,demonstrating superior generalizability and robustness.The proposed ML framework offers an efficient,accurate,and data-driven alternative to traditional methods for estimating caisson capacity in stratified soils.This approach can aid in reducing computational costs while improving reliability in the early stages of foundation design.展开更多
Machine-learning is a robust technique for understanding pollution characteristics of surface ozone,which are at high levels in urban China.This study introduced an innovative approach combining trend decomposition wi...Machine-learning is a robust technique for understanding pollution characteristics of surface ozone,which are at high levels in urban China.This study introduced an innovative approach combining trend decomposition with Random Forest algorithm to investigate ozone dynamics and formation regimes in a coastal area of China.During the period of 2017–2022,significant inter-annual fluctuations emerged,with peaks in mid-2017 attributed to volatile organic compounds(VOCs),and in late-2019 influenced by air temperature.Multifaceted periodicities(daily,weekly,holiday,and yearly)in ozone were revealed,elucidating substantial influences of daily and yearly components on ozone periodicity.A VOC-sensitive ozone formation regime was identified,characterized by lower VOCs/NO_(x) ratios(average=0.88)and significant positive correlations between ozone and VOCs.This interplay manifested in elevated ozone duringweekends,holidays,and pandemic lockdowns.Key variables influencing ozone across diverse timescaleswere uncovered,with solar radiation and temperature driving daily and yearly ozone variations,respectively.Precursor substances,particularly VOCs,significantly shaped weekly/holiday patterns and long-term trends of ozone.Specifically,acetone,ethane,hexanal,and toluene had a notable impact on the multi-year ozone trend,emphasizing the urgency of VOC regulation.Furthermore,our observations indicated that NO_(x) primarily drived the stochastic variations in ozone,a distinguishing characteristic of regions with heavy traffic.This research provides novel insights into ozone dynamics in coastal urban areas and highlights the importance of integrating statistical and machinelearning methods in atmospheric pollution studies,with implications for targeted mitigation strategies beyond this specific region and pollutant.展开更多
Multivariate time series forecasting iswidely used in traffic planning,weather forecasting,and energy consumption.Series decomposition algorithms can help models better understand the underlying patterns of the origin...Multivariate time series forecasting iswidely used in traffic planning,weather forecasting,and energy consumption.Series decomposition algorithms can help models better understand the underlying patterns of the original series to improve the forecasting accuracy of multivariate time series.However,the decomposition kernel of previous decomposition-based models is fixed,and these models have not considered the differences in frequency fluctuations between components.These problems make it difficult to analyze the intricate temporal variations of real-world time series.In this paper,we propose a series decomposition-based Mamba model,DecMamba,to obtain the intricate temporal dependencies and the dependencies among different variables of multivariate time series.A variable-level adaptive kernel combination search module is designed to interact with information on different trends and periods between variables.Two backbone structures are proposed to emphasize the differences in frequency fluctuations of seasonal and trend components.Mamba with superior performance is used instead of a Transformer in backbone structures to capture the dependencies among different variables.A new embedding block is designed to capture the temporal features better,especially for the high-frequency seasonal component whose semantic information is difficult to acquire.A gating mechanism is introduced to the decoder in the seasonal backbone to improve the prediction accuracy.A comparison with ten state-of-the-art models on seven real-world datasets demonstrates that DecMamba can better model the temporal dependencies and the dependencies among different variables,guaranteeing better prediction performance for multivariate time series.展开更多
Fluoride-based electrolyte exhibits extraordinarily high oxidative stability in high-voltage lithium metal batteries(h-LMBs) due to the inherent low highest occupied molecular orbital(HOMO) of fiuorinated solvents. Ho...Fluoride-based electrolyte exhibits extraordinarily high oxidative stability in high-voltage lithium metal batteries(h-LMBs) due to the inherent low highest occupied molecular orbital(HOMO) of fiuorinated solvents. However, such fascinating properties do not bring long-term cyclability of h-LMBs. One of critical challenges is the interface instability in contacting with the Li metal anode, as fiuorinated solvents are highly susceptible to exceptionally reductive metallic Li attributed to its low lowest unoccupied molecular orbital(LUMO), which leads to significant consumption of the fiuorinated components upon cycling.Herein, attenuating reductive decomposition of fiuorinated electrolytes is proposed to circumvent rapid electrolyte consumption. Specifically, the vinylene carbonate(VC) is selected to tame the reduction decomposition by preferentially forming protective layer on the Li anode. This work, experimentally and computationally, demonstrates the importance of pre-passivation of Li metal anodes at high voltage to attenuate the decomposition of fiuoroethylene carbonate(FEC). It is expected to enrich the understanding of how VC attenuate the reactivity of FEC, thereby extending the cycle life of fiuorinated electrolytes in high-voltage Li-metal batteries.展开更多
The multi-objective particle swarm optimization algorithm(MOPSO)is widely used to solve multi-objective optimization problems.In the article,amulti-objective particle swarm optimization algorithmbased on decomposition...The multi-objective particle swarm optimization algorithm(MOPSO)is widely used to solve multi-objective optimization problems.In the article,amulti-objective particle swarm optimization algorithmbased on decomposition and multi-selection strategy is proposed to improve the search efficiency.First,two update strategies based on decomposition are used to update the evolving population and external archive,respectively.Second,a multiselection strategy is designed.The first strategy is for the subspace without a non-dominated solution.Among the neighbor particles,the particle with the smallest penalty-based boundary intersection value is selected as the global optimal solution and the particle far away fromthe search particle and the global optimal solution is selected as the personal optimal solution to enhance global search.The second strategy is for the subspace with a non-dominated solution.In the neighbor particles,two particles are randomly selected,one as the global optimal solution and the other as the personal optimal solution,to enhance local search.The third strategy is for Pareto optimal front(PF)discontinuity,which is identified by the cumulative number of iterations of the subspace without non-dominated solutions.In the subsequent iteration,a new probability distribution is used to select from the remaining subspaces to search.Third,an adaptive inertia weight update strategy based on the dominated degree is designed to further improve the search efficiency.Finally,the proposed algorithmis compared with fivemulti-objective particle swarm optimization algorithms and five multi-objective evolutionary algorithms on 22 test problems.The results show that the proposed algorithm has better performance.展开更多
In this paper,we prove that L(K_(x,y))(λ),theλ-fold line graph of the complete bipartite graph Ka,y,has a C_(6)-decomposition if and only if ry≥6,λxy(c+y-2)=0(mod 12)and(x+y)=0(mod 2),where x,y are nonnegative int...In this paper,we prove that L(K_(x,y))(λ),theλ-fold line graph of the complete bipartite graph Ka,y,has a C_(6)-decomposition if and only if ry≥6,λxy(c+y-2)=0(mod 12)and(x+y)=0(mod 2),where x,y are nonnegative integers and(x,y)≠(2,4)or(2,5).展开更多
To address the issues of peak overlap caused by complex matrices in agricultural product terahertz(THz)spectral signals and the dynamic,nonlinear interference induced by environmental and system noise,this study explo...To address the issues of peak overlap caused by complex matrices in agricultural product terahertz(THz)spectral signals and the dynamic,nonlinear interference induced by environmental and system noise,this study explores the feasibility of adaptive-signal-decomposition-based denoising methods to improve THz spectral quality.THz time-domain spectroscopy(THz-TDS)combined with an attenuated total reflection(ATR)accessory was used to collect THz absorbance spectra from 48 peanut samples.Taking the quantitative prediction model of peanut moisture content based on THz-ATR as an example,wavelet transform(WT),empirical mode decomposition(EMD),local mean decomposition(LMD),and its improved methods-segmented local mean decomposition(SLMD)and piecewise mirror extension local mean decomposition(PME-LMD)-were employed for spectral denoising.The applicability of different denoising methods was evaluated using a support vector regression(SVR)model.Experimental results show that the peanut moisture content prediction model constructed after PME-LMD denoising achieved the best performance,with a root mean square error(RMSE),coefficient of determination(R^(2)),and mean absolute percentage error(MAPE)of 0.010,0.912,and 0.040,respectively.Compared with traditional methods,PME-LMD significantly improved spectral quality and model prediction performance.The PME-LMD denoising strategy proposed in this study effectively suppresses non-uniform noise interference in THz spectral signals,providing an efficient and accurate preprocessing method for THz spectral analysis of agricultural products.This research provides theoretical support and technical guidance for the application of THz technology for detecting agricultural product quality.展开更多
In recent years,ozone has become one of the key pollutants affecting the urban air qual-ity.Direct catalytic decomposition of ozone emerges as an effective method for ozone re-moval.Field experimentswere conducted to ...In recent years,ozone has become one of the key pollutants affecting the urban air qual-ity.Direct catalytic decomposition of ozone emerges as an effective method for ozone re-moval.Field experimentswere conducted to evaluate the effectiveness of exteriorwall coat-ings with ozone decomposition catalysts for ozone removal in practical applications.ANSYS 2020R1 software was first used for simulation and analysis of ozone concentration and flow fields to investigate the decomposition boundary of these wall coatings.The results show that the exterior wall coatings with manganese-based catalysts can effectively reduce the ozone concentration near the wall coating.The ozone decomposition efficiency is nega-tively correlated with the distance fromthe coating and the decomposition boundary range is around 18 m.The decomposition boundary will increase with the increase of tempera-ture,and decrease with the increase of the wind speed and the relative humidity.These results underscore the viability of using exterior wall coatings with catalysts for controlling ozone pollution in atmospheric environments.This approach presents a promising avenue for addressing ozone pollution through self-purifying materials on building external wall.展开更多
In the realm of nonlinear integrable systems,the presence of decompositions facilitates the establishment of linear superposition solutions and the derivation of novel coupled systems exhibiting nonlinear integrabilit...In the realm of nonlinear integrable systems,the presence of decompositions facilitates the establishment of linear superposition solutions and the derivation of novel coupled systems exhibiting nonlinear integrability.By focusing on single-component decompositions within the potential BKP hierarchy,it has been observed that specific linear superpositions of decomposition solutions remain consistent with the underlying equations.Moreover,through the implementation of multi-component decompositions within the potential BKP hierarchy,successful endeavors have been undertaken to formulate linear superposition solutions and novel coupled Kd V-type systems that resist decoupling via alterations in dependent variables.展开更多
This study investigates the effects of radiation force due to the rotational pitch motion of a wave energy device,which comprises a coaxial bottom-mounted cylindrical caisson in a two-layer fluid,along with a submerge...This study investigates the effects of radiation force due to the rotational pitch motion of a wave energy device,which comprises a coaxial bottom-mounted cylindrical caisson in a two-layer fluid,along with a submerged cylindrical buoy.The system is modeled as a two-layer fluid with infinite horizontal extent and finite depth.The radiation problem is analyzed in the context of linear water waves.The fluid domain is divided into outer and inner zones,and mathematical solutions for the pitch radiating potential are derived for the corresponding boundary valve problem in these zones using the separation of variables approach.Using the matching eigenfunction expansion method,the unknown coefficients in the analytical expression of the radiation potentials are evaluated.The resulting radiation potential is then used to compute the added mass and damping coefficients.Several numerical results for the added mass and damping coefficients are investigated for numerous parameters,particularly the effects of the cylinder radius,the draft of the submerged cylinder,and the density proportion between the two fluid layers across different frequency ranges.The major findings are presented and discussed.展开更多
Network Intrusion Detection System(NIDS)detection of minority class attacks is always a difficult task when dealing with attacks in complex network environments.To improve the detection capability of minority-class at...Network Intrusion Detection System(NIDS)detection of minority class attacks is always a difficult task when dealing with attacks in complex network environments.To improve the detection capability of minority-class attacks,this study proposes an intrusion detection method based on a two-layer structure.The first layer employs a CNN-BiLSTM model incorporating an attention mechanism to classify network traffic into normal traffic,majority class attacks,and merged minority class attacks.The second layer further segments the minority class attacks through Stacking ensemble learning.The datasets are selected from the generic network dataset CIC-IDS2017,NSL-KDD,and the industrial network dataset Mississippi Gas Pipeline dataset to enhance the generalization and practical applicability of the model.Experimental results show that the proposed model achieves an overall detection accuracy of 99%,99%,and 95%on the CIC-IDS2017,NSL-KDD,and industrial network datasets,respectively.It also significantly outperforms traditional methods in terms of detection accuracy and recall rate for minority class attacks.Compared with the single-layer deep learning model,the two-layer structure effectively reduces the false alarm rate while improving the minority-class attack detection performance.The research in this paper not only improves the adaptability of NIDS to complex network environments but also provides a new solution for minority-class attack detection in industrial network security.展开更多
We presented the preparation and analysis of La_(1-x)K_(x)CoO_(3)(x=0.1-0.4)catalysts,supported on microwave-absorbing ceramic carriers,using the sol-gel method.We systematically investigated the effects of various re...We presented the preparation and analysis of La_(1-x)K_(x)CoO_(3)(x=0.1-0.4)catalysts,supported on microwave-absorbing ceramic carriers,using the sol-gel method.We systematically investigated the effects of various reaction conditions under microwave irradiation(0-50 W).These conditions included reaction temperatures(300-600℃),oxygen concentrations(0-6%),and varying K^(+)doping levels on the catalysts'activity.The crystalline phase,microstructure,and the catalytic activity of the catalyst were analyzed by XRD,TEM,H_2-TPR,and O_(2)-TPD.The experimental results reveal that La_(1-x)K_(x)CoO_(3)(x=0.1-0.4)catalysts consistently form homogeneous perovskite nanoparticles across different doping levels.The NO decomposition efficiency on these catalysts initially increases and then decreases with variations in doping amount,temperature,and microwave power.Additionally,an increase in oxygen concentration positively influences NO conversion rates.The optimal performance is observed with La_(0.7)K_(0.3)CoO_(3)catalyst under conditions of x=0.3,400℃,10 W microwave power,and 4%oxygen concentration,achieving a peak NO conversion rate of La_(0.7)K_(0.3)CoO_(3)catalyst is 93.1%.展开更多
As an independent thermodynamic parameter,pressure significantly influences interatomic distances,leading to an increase in material density.In this work,we employ the CALYPSO structure search and density functional t...As an independent thermodynamic parameter,pressure significantly influences interatomic distances,leading to an increase in material density.In this work,we employ the CALYPSO structure search and density functional theory calculations to explore the structural phase transitions and electronic properties of calcium-sulfur compounds(Ca_(x)S_(1-x),where x=1/4,1/3,1/2,2/3,3/4,4/5)under 0-1200 GPa.The calculated formation enthalpies suggest that Ca_(x)S_(1-x)compounds undergo multiple phase transitions and eventually decompose into elemental Ca and S,challenging the traditional view that pressure stabilizes and densifies compounds.The analysis of formation enthalpy indicates that an increase in pressure leads to a rise in internal energy and the PV term,resulting in thermodynamic instability.Bader charge analysis reveals that this phenomenon is attributed to a decrease in charge transfer under high pressure.The activation of Ca-3d orbitals is significantly enhanced under pressure,leading to competition with Ca-4s orbitals and S-3p orbitals.This may cause the formation enthalpy minimum on the convex hull to shift sequentially from CaS to CaS_(3),then to Ca_(3)S and Ca_(2)S,and finally back to CaS.These findings provide critical insights into the behavior of alkaline-earth metal sulfides under high pressure,with implications for the synthesis and application of novel materials under extreme conditions and for understanding element distribution in planetary interiors.展开更多
BACKGROUND The early diagnosis rate of pancreatic ductal adenocarcinoma(PDAC)is low and the prognosis is poor.It is important to develop an interpretable noninvasive early diagnostic model in clinical practice.AIM To ...BACKGROUND The early diagnosis rate of pancreatic ductal adenocarcinoma(PDAC)is low and the prognosis is poor.It is important to develop an interpretable noninvasive early diagnostic model in clinical practice.AIM To develop an interpretable noninvasive early diagnostic model for PDAC using plasma extracellular vesicle long RNA(EvlRNA).METHODS The diagnostic model was constructed based on plasma EvlRNA data.During the process of establishing the model,EvlRNA-index was introduced,and four algorithms were adopted to calculate EvlRNA-index.After the model was successfully constructed,performance evaluation was conducted.A series of bioinformatics methods were adopted to explore the potential mechanism of EvlRNA-index as the input feature of the model.And the relationship between key characteristics and PDAC were explored at the single-cell level.RESULTS A novel interpretable machine learning framework was developed based on plasma EvlRNA.In this framework,a two-layer classifier was established.A new concept was proposed:EvlRNA-index.Based on EvlRNA-index,a cancer diagnostic model was established,and a good diagnostic effect was achieved.The accuracy of PDACandCPvsHealth-Probabilistic PCA Index-SVM(PDAC and chronic pancreatitis vs health-probabilistic principal component analysis index-support vector machine)(1-18)was 91.51%,with Mathew’s correlation coefficient 0.7760 and area under the curve 0.9560.In the second layer of the model,the accuracy of PDACvsCP-Probabilistic PCA Index-RF(PDAC vs chronic pancreatitis-probabilistic principal component analysis index-random forest)(2-17)was 93.83%,with Mathew’s correlation coefficient 0.8422 and area under the curve 0.9698.Forty-nine PDAC-related genes were identified,among which 16 were known,inferring that the remaining ones were also PDAC-related genes.CONCLUSION An interpretable two-layer machine learning framework was proposed for early diagnosis and prediction of PDAC based on plasma EvlRNA,providing new insights into the clinical value of EvlRNA.展开更多
Organic matter(OM)derived from the decomposition of crop residues plays a key role as a sorbent for cadmium(Cd)immobilization.Few studies have explored the straw decomposition processes with the presence of minerals,a...Organic matter(OM)derived from the decomposition of crop residues plays a key role as a sorbent for cadmium(Cd)immobilization.Few studies have explored the straw decomposition processes with the presence of minerals,and the effect of newly generated organomineral complexes on heavy metal adsorption.In this study,we investigated the variations in structure and composition during the rice straw decomposition with or without minerals(goethite and kaolinite),as well as the adsorption behavior and mechanisms by which straw decomposition affects Cd immobilization.The degree of humification of extracted straw organic matter was assessed using excitation-emission matrix(EEM)fluorescence and Ultraviolet-visible spectroscopy(UV-vis),while employing FTIR spectroscopy and XPS to characterize the adsorption mechanisms.The spectra analysis revealed the enrichment of highly aromatic and hydrophobic components,indicating that the degree of straw decomposition and humification were further intensified during incubation.Additionally,the existence of goethite(SG)accelerated the humification of OM.Sorption experiments revealed that the straw humification increased Cd adsorption capacity.Notably,SG exhibited significantly higher adsorption performance compared to the organic matter without minerals(RS)and the existence of kaolinite(SK).Further analysis using FT-IR spectroscopy and XPS verified that the primarymechanisms involved in Cd immobilization were complexion with—OH and—COOH,as well as the formation of Cd-πbinds with aromatic C=C on the surface of solid OMs.These findings will facilitate understanding the interactions of the rice straw decomposing with soil minerals and its remediation effect on Cd-contaminated farmland.展开更多
Earthquakes can cause significant hazards,leading to loss of human life and property damage.In this study,we applied the masking empirical mode decomposition(EMD)to the Gyeongju earthquake(ML 5.8)in South Korea and th...Earthquakes can cause significant hazards,leading to loss of human life and property damage.In this study,we applied the masking empirical mode decomposition(EMD)to the Gyeongju earthquake(ML 5.8)in South Korea and the Abe Pura earthquake(ML 7.0)in Indonesia and then revealed that the 7th intrinsic mode function(IMF7),composed of waveforms with a period band of approximately 150 to 750 minutes.展开更多
In the wake of major natural disasters or human-made disasters,the communication infrastruc-ture within disaster-stricken areas is frequently dam-aged.Unmanned aerial vehicles(UAVs),thanks to their merits such as rapi...In the wake of major natural disasters or human-made disasters,the communication infrastruc-ture within disaster-stricken areas is frequently dam-aged.Unmanned aerial vehicles(UAVs),thanks to their merits such as rapid deployment and high mobil-ity,are commonly regarded as an ideal option for con-structing temporary communication networks.Con-sidering the limited computing capability and battery power of UAVs,this paper proposes a two-layer UAV cooperative computing offloading strategy for emer-gency disaster relief scenarios.The multi-agent twin delayed deep deterministic policy gradient(MATD3)algorithm integrated with prioritized experience replay(PER)is utilized to jointly optimize the scheduling strategies of UAVs,task offloading ratios,and their mobility,aiming to diminish the energy consumption and delay of the system to the minimum.In order to address the aforementioned non-convex optimiza-tion issue,a Markov decision process(MDP)has been established.The results of simulation experiments demonstrate that,compared with the other four base-line algorithms,the algorithm introduced in this paper exhibits better convergence performance,verifying its feasibility and efficacy.展开更多
Formalizing complex processes and phenomena of a real-world problem may require a large number of variables and constraints,resulting in what is termed a large-scale optimization problem.Nowadays,such large-scale opti...Formalizing complex processes and phenomena of a real-world problem may require a large number of variables and constraints,resulting in what is termed a large-scale optimization problem.Nowadays,such large-scale optimization problems are solved using computing machines,leading to an enormous computational time being required,which may delay deriving timely solutions.Decomposition methods,which partition a large-scale optimization problem into lower-dimensional subproblems,represent a key approach to addressing time-efficiency issues.There has been significant progress in both applied mathematics and emerging artificial intelligence approaches on this front.This work aims at providing an overview of the decomposition methods from both the mathematics and computer science points of view.We also remark on the state-of-the-art developments and recent applications of the decomposition methods,and discuss the future research and development perspectives.展开更多
Heterogeneous graphs generally refer to graphs with different types of nodes and edges.A common approach for extracting useful information from heterogeneous graphs is to use meta-graphs,which can be seen as a special...Heterogeneous graphs generally refer to graphs with different types of nodes and edges.A common approach for extracting useful information from heterogeneous graphs is to use meta-graphs,which can be seen as a special kind of directed acyclic graph with same node and edge types as the heterogeneous graph.However,how to design proper metagraphs is challenging.Recently,there have been many works on learning suitable metagraphs from a heterogeneous graph.Existing methods generally introduce continuous weights for edges that are independent of each other,which ignores the topological structures of meta-graphs and can be ineffective.To address this issue,the authors propose a new viewpoint from tensor on learning meta-graphs.Such a viewpoint not only helps interpret the limitation of existing works by CANDECOMP/PARAFAC(CP)decomposition,but also inspires us to propose a topology-aware tensor decomposition,called TENSUS,that reflects the structure of DAGs.The proposed topology-aware tensor decomposition is easy to use and simple to implement,and it can be taken as a plug-in part to upgrade many existing works,including node classification and recommendation on heterogeneous graphs.Experimental results on different tasks demonstrate that the proposed method can significantly improve the state-of-the-arts for all these tasks.展开更多
基金financially supported by the National Natural Science Foundation of China(Nos.52034002 and U2202254)the Fundamental Research Funds for the Central Universities,China(No.FRF-TT-19-001)。
文摘The sulfation and decomposition process has proven effective in selectively extracting lithium from lepidolite.It is essential to clarify the thermochemical behavior and kinetic parameters of decomposition reactions.Accordingly,comprehensive kinetic study by employing thermalgravimetric analysis at various heating rates was presented in this paper.Two main weight loss regions were observed during heating.The initial region corresponded to the dehydration of crystal water,whereas the subsequent region with overlapping peaks involved complex decomposition reactions.The overlapping peaks were separated into two individual reaction peaks and the activation energy of each peak was calculated using isoconversional kinetics methods.The activation energy of peak 1 exhibited a continual increase as the reaction conversion progressed,while that of peak 2 steadily decreased.The optimal kinetic models,identified as belonging to the random nucleation and subsequent growth category,provided valuable insights into the mechanism of the decomposition reactions.Furthermore,the adjustment factor was introduced to reconstruct the kinetic mechanism models,and the reconstructed models described the kinetic mechanism model more accurately for the decomposition reactions.This study enhanced the understanding of the thermochemical behavior and kinetic parameters of the lepidolite sulfation product decomposition reactions,further providing theoretical basis for promoting the selective extraction of lithium.
文摘Open caissons are widely used in foundation engineering because of their load-bearing efficiency and adaptability in diverse soil conditions.However,accurately predicting their undrained bearing capacity in layered soils remains a complex challenge.This study presents a novel application of five ensemble machine(ML)algorithms-random forest(RF),gradient boosting machine(GBM),extreme gradient boosting(XGBoost),adaptive boosting(AdaBoost),and categorical boosting(CatBoost)-to predict the undrained bearing capacity factor(Nc)of circular open caissons embedded in two-layered clay on the basis of results from finite element limit analysis(FELA).The input dataset consists of 1188 numerical simulations using the Tresca failure criterion,varying in geometrical and soil parameters.The FELA was performed via OptumG2 software with adaptive meshing techniques and verified against existing benchmark studies.The ML models were trained on 70% of the dataset and tested on the remaining 30%.Their performance was evaluated using six statistical metrics:coefficient of determination(R²),mean absolute error(MAE),root mean squared error(RMSE),index of scatter(IOS),RMSE-to-standard deviation ratio(RSR),and variance explained factor(VAF).The results indicate that all the models achieved high accuracy,with R²values exceeding 97.6%and RMSE values below 0.02.Among them,AdaBoost and CatBoost consistently outperformed the other methods across both the training and testing datasets,demonstrating superior generalizability and robustness.The proposed ML framework offers an efficient,accurate,and data-driven alternative to traditional methods for estimating caisson capacity in stratified soils.This approach can aid in reducing computational costs while improving reliability in the early stages of foundation design.
基金supported by Ningbo Natural Science Foundation(No.2023J059)Ningbo Commonweal Programme Key Project(No.2023S038)Guangxi Key Research and Development Programme(No.GuikeAB21220063).
文摘Machine-learning is a robust technique for understanding pollution characteristics of surface ozone,which are at high levels in urban China.This study introduced an innovative approach combining trend decomposition with Random Forest algorithm to investigate ozone dynamics and formation regimes in a coastal area of China.During the period of 2017–2022,significant inter-annual fluctuations emerged,with peaks in mid-2017 attributed to volatile organic compounds(VOCs),and in late-2019 influenced by air temperature.Multifaceted periodicities(daily,weekly,holiday,and yearly)in ozone were revealed,elucidating substantial influences of daily and yearly components on ozone periodicity.A VOC-sensitive ozone formation regime was identified,characterized by lower VOCs/NO_(x) ratios(average=0.88)and significant positive correlations between ozone and VOCs.This interplay manifested in elevated ozone duringweekends,holidays,and pandemic lockdowns.Key variables influencing ozone across diverse timescaleswere uncovered,with solar radiation and temperature driving daily and yearly ozone variations,respectively.Precursor substances,particularly VOCs,significantly shaped weekly/holiday patterns and long-term trends of ozone.Specifically,acetone,ethane,hexanal,and toluene had a notable impact on the multi-year ozone trend,emphasizing the urgency of VOC regulation.Furthermore,our observations indicated that NO_(x) primarily drived the stochastic variations in ozone,a distinguishing characteristic of regions with heavy traffic.This research provides novel insights into ozone dynamics in coastal urban areas and highlights the importance of integrating statistical and machinelearning methods in atmospheric pollution studies,with implications for targeted mitigation strategies beyond this specific region and pollutant.
基金supported in part by the Interdisciplinary Project of Dalian University(DLUXK-2023-ZD-001).
文摘Multivariate time series forecasting iswidely used in traffic planning,weather forecasting,and energy consumption.Series decomposition algorithms can help models better understand the underlying patterns of the original series to improve the forecasting accuracy of multivariate time series.However,the decomposition kernel of previous decomposition-based models is fixed,and these models have not considered the differences in frequency fluctuations between components.These problems make it difficult to analyze the intricate temporal variations of real-world time series.In this paper,we propose a series decomposition-based Mamba model,DecMamba,to obtain the intricate temporal dependencies and the dependencies among different variables of multivariate time series.A variable-level adaptive kernel combination search module is designed to interact with information on different trends and periods between variables.Two backbone structures are proposed to emphasize the differences in frequency fluctuations of seasonal and trend components.Mamba with superior performance is used instead of a Transformer in backbone structures to capture the dependencies among different variables.A new embedding block is designed to capture the temporal features better,especially for the high-frequency seasonal component whose semantic information is difficult to acquire.A gating mechanism is introduced to the decoder in the seasonal backbone to improve the prediction accuracy.A comparison with ten state-of-the-art models on seven real-world datasets demonstrates that DecMamba can better model the temporal dependencies and the dependencies among different variables,guaranteeing better prediction performance for multivariate time series.
基金supported by the National Natural Science Foundation of China (Nos. 22379121, 62005216)Basic Public Welfare Research Program of Zhejiang (No. LQ22F050013)+1 种基金Zhejiang Province Key Laboratory of Flexible Electronics Open Fund (2023FE005)Shenzhen Foundation Research Program (No. JCYJ20220530112812028)。
文摘Fluoride-based electrolyte exhibits extraordinarily high oxidative stability in high-voltage lithium metal batteries(h-LMBs) due to the inherent low highest occupied molecular orbital(HOMO) of fiuorinated solvents. However, such fascinating properties do not bring long-term cyclability of h-LMBs. One of critical challenges is the interface instability in contacting with the Li metal anode, as fiuorinated solvents are highly susceptible to exceptionally reductive metallic Li attributed to its low lowest unoccupied molecular orbital(LUMO), which leads to significant consumption of the fiuorinated components upon cycling.Herein, attenuating reductive decomposition of fiuorinated electrolytes is proposed to circumvent rapid electrolyte consumption. Specifically, the vinylene carbonate(VC) is selected to tame the reduction decomposition by preferentially forming protective layer on the Li anode. This work, experimentally and computationally, demonstrates the importance of pre-passivation of Li metal anodes at high voltage to attenuate the decomposition of fiuoroethylene carbonate(FEC). It is expected to enrich the understanding of how VC attenuate the reactivity of FEC, thereby extending the cycle life of fiuorinated electrolytes in high-voltage Li-metal batteries.
基金supported by National Natural Science Foundations of China(nos.12271326,62102304,61806120,61502290,61672334,61673251)China Postdoctoral Science Foundation(no.2015M582606)+2 种基金Industrial Research Project of Science and Technology in Shaanxi Province(nos.2015GY016,2017JQ6063)Fundamental Research Fund for the Central Universities(no.GK202003071)Natural Science Basic Research Plan in Shaanxi Province of China(no.2022JM-354).
文摘The multi-objective particle swarm optimization algorithm(MOPSO)is widely used to solve multi-objective optimization problems.In the article,amulti-objective particle swarm optimization algorithmbased on decomposition and multi-selection strategy is proposed to improve the search efficiency.First,two update strategies based on decomposition are used to update the evolving population and external archive,respectively.Second,a multiselection strategy is designed.The first strategy is for the subspace without a non-dominated solution.Among the neighbor particles,the particle with the smallest penalty-based boundary intersection value is selected as the global optimal solution and the particle far away fromthe search particle and the global optimal solution is selected as the personal optimal solution to enhance global search.The second strategy is for the subspace with a non-dominated solution.In the neighbor particles,two particles are randomly selected,one as the global optimal solution and the other as the personal optimal solution,to enhance local search.The third strategy is for Pareto optimal front(PF)discontinuity,which is identified by the cumulative number of iterations of the subspace without non-dominated solutions.In the subsequent iteration,a new probability distribution is used to select from the remaining subspaces to search.Third,an adaptive inertia weight update strategy based on the dominated degree is designed to further improve the search efficiency.Finally,the proposed algorithmis compared with fivemulti-objective particle swarm optimization algorithms and five multi-objective evolutionary algorithms on 22 test problems.The results show that the proposed algorithm has better performance.
文摘In this paper,we prove that L(K_(x,y))(λ),theλ-fold line graph of the complete bipartite graph Ka,y,has a C_(6)-decomposition if and only if ry≥6,λxy(c+y-2)=0(mod 12)and(x+y)=0(mod 2),where x,y are nonnegative integers and(x,y)≠(2,4)or(2,5).
基金Supported by the National Key R&D Program of China(2023YFD2101001)National Natural Science Foundation of China(32202144,61807001)。
文摘To address the issues of peak overlap caused by complex matrices in agricultural product terahertz(THz)spectral signals and the dynamic,nonlinear interference induced by environmental and system noise,this study explores the feasibility of adaptive-signal-decomposition-based denoising methods to improve THz spectral quality.THz time-domain spectroscopy(THz-TDS)combined with an attenuated total reflection(ATR)accessory was used to collect THz absorbance spectra from 48 peanut samples.Taking the quantitative prediction model of peanut moisture content based on THz-ATR as an example,wavelet transform(WT),empirical mode decomposition(EMD),local mean decomposition(LMD),and its improved methods-segmented local mean decomposition(SLMD)and piecewise mirror extension local mean decomposition(PME-LMD)-were employed for spectral denoising.The applicability of different denoising methods was evaluated using a support vector regression(SVR)model.Experimental results show that the peanut moisture content prediction model constructed after PME-LMD denoising achieved the best performance,with a root mean square error(RMSE),coefficient of determination(R^(2)),and mean absolute percentage error(MAPE)of 0.010,0.912,and 0.040,respectively.Compared with traditional methods,PME-LMD significantly improved spectral quality and model prediction performance.The PME-LMD denoising strategy proposed in this study effectively suppresses non-uniform noise interference in THz spectral signals,providing an efficient and accurate preprocessing method for THz spectral analysis of agricultural products.This research provides theoretical support and technical guidance for the application of THz technology for detecting agricultural product quality.
基金supported by the National Natural Science Foundation of China(Nos.52470114 and 52022104)the National Key R&D Program of China(No.2022YFC3702802)the Youth Innovation Promotion Association,CAS(No.Y2021020).
文摘In recent years,ozone has become one of the key pollutants affecting the urban air qual-ity.Direct catalytic decomposition of ozone emerges as an effective method for ozone re-moval.Field experimentswere conducted to evaluate the effectiveness of exteriorwall coat-ings with ozone decomposition catalysts for ozone removal in practical applications.ANSYS 2020R1 software was first used for simulation and analysis of ozone concentration and flow fields to investigate the decomposition boundary of these wall coatings.The results show that the exterior wall coatings with manganese-based catalysts can effectively reduce the ozone concentration near the wall coating.The ozone decomposition efficiency is nega-tively correlated with the distance fromthe coating and the decomposition boundary range is around 18 m.The decomposition boundary will increase with the increase of tempera-ture,and decrease with the increase of the wind speed and the relative humidity.These results underscore the viability of using exterior wall coatings with catalysts for controlling ozone pollution in atmospheric environments.This approach presents a promising avenue for addressing ozone pollution through self-purifying materials on building external wall.
基金sponsored by the National Natural Science Foundations of China under Grant Nos.12301315,12235007,11975131the Zhejiang Provincial Natural Science Foundation of China under Grant No.LQ20A010009。
文摘In the realm of nonlinear integrable systems,the presence of decompositions facilitates the establishment of linear superposition solutions and the derivation of novel coupled systems exhibiting nonlinear integrability.By focusing on single-component decompositions within the potential BKP hierarchy,it has been observed that specific linear superpositions of decomposition solutions remain consistent with the underlying equations.Moreover,through the implementation of multi-component decompositions within the potential BKP hierarchy,successful endeavors have been undertaken to formulate linear superposition solutions and novel coupled Kd V-type systems that resist decoupling via alterations in dependent variables.
基金supported by MHRD as researcher C.K.Neog received the MHRD Institute GATE scholarship from Govt.of India.
文摘This study investigates the effects of radiation force due to the rotational pitch motion of a wave energy device,which comprises a coaxial bottom-mounted cylindrical caisson in a two-layer fluid,along with a submerged cylindrical buoy.The system is modeled as a two-layer fluid with infinite horizontal extent and finite depth.The radiation problem is analyzed in the context of linear water waves.The fluid domain is divided into outer and inner zones,and mathematical solutions for the pitch radiating potential are derived for the corresponding boundary valve problem in these zones using the separation of variables approach.Using the matching eigenfunction expansion method,the unknown coefficients in the analytical expression of the radiation potentials are evaluated.The resulting radiation potential is then used to compute the added mass and damping coefficients.Several numerical results for the added mass and damping coefficients are investigated for numerous parameters,particularly the effects of the cylinder radius,the draft of the submerged cylinder,and the density proportion between the two fluid layers across different frequency ranges.The major findings are presented and discussed.
基金supported by the Institute of Information&Communications Technology Planning&Evaluation(IITP)—Innovative Human Resource Development for Local Intellectualization program grant funded by the Korea government(MSIT)(IITP-2025-RS-2022-00156334)in part by Liaoning Province Nature Fund Project(2024-BSLH-214).
文摘Network Intrusion Detection System(NIDS)detection of minority class attacks is always a difficult task when dealing with attacks in complex network environments.To improve the detection capability of minority-class attacks,this study proposes an intrusion detection method based on a two-layer structure.The first layer employs a CNN-BiLSTM model incorporating an attention mechanism to classify network traffic into normal traffic,majority class attacks,and merged minority class attacks.The second layer further segments the minority class attacks through Stacking ensemble learning.The datasets are selected from the generic network dataset CIC-IDS2017,NSL-KDD,and the industrial network dataset Mississippi Gas Pipeline dataset to enhance the generalization and practical applicability of the model.Experimental results show that the proposed model achieves an overall detection accuracy of 99%,99%,and 95%on the CIC-IDS2017,NSL-KDD,and industrial network datasets,respectively.It also significantly outperforms traditional methods in terms of detection accuracy and recall rate for minority class attacks.Compared with the single-layer deep learning model,the two-layer structure effectively reduces the false alarm rate while improving the minority-class attack detection performance.The research in this paper not only improves the adaptability of NIDS to complex network environments but also provides a new solution for minority-class attack detection in industrial network security.
文摘We presented the preparation and analysis of La_(1-x)K_(x)CoO_(3)(x=0.1-0.4)catalysts,supported on microwave-absorbing ceramic carriers,using the sol-gel method.We systematically investigated the effects of various reaction conditions under microwave irradiation(0-50 W).These conditions included reaction temperatures(300-600℃),oxygen concentrations(0-6%),and varying K^(+)doping levels on the catalysts'activity.The crystalline phase,microstructure,and the catalytic activity of the catalyst were analyzed by XRD,TEM,H_2-TPR,and O_(2)-TPD.The experimental results reveal that La_(1-x)K_(x)CoO_(3)(x=0.1-0.4)catalysts consistently form homogeneous perovskite nanoparticles across different doping levels.The NO decomposition efficiency on these catalysts initially increases and then decreases with variations in doping amount,temperature,and microwave power.Additionally,an increase in oxygen concentration positively influences NO conversion rates.The optimal performance is observed with La_(0.7)K_(0.3)CoO_(3)catalyst under conditions of x=0.3,400℃,10 W microwave power,and 4%oxygen concentration,achieving a peak NO conversion rate of La_(0.7)K_(0.3)CoO_(3)catalyst is 93.1%.
基金Project supported by the National Natural Science Foundation of China(Grant Nos.11974154 and 12304278)the Taishan Scholars Special Funding for Construction Projects(Grant No.tstp20230622)+1 种基金the Natural Science Foundation of Shandong Province(Grant Nos.ZR2022MA004,ZR2023QA127,and ZR2024QA121)the Special Foundation of Yantai for Leading Talents above Provincial Level.
文摘As an independent thermodynamic parameter,pressure significantly influences interatomic distances,leading to an increase in material density.In this work,we employ the CALYPSO structure search and density functional theory calculations to explore the structural phase transitions and electronic properties of calcium-sulfur compounds(Ca_(x)S_(1-x),where x=1/4,1/3,1/2,2/3,3/4,4/5)under 0-1200 GPa.The calculated formation enthalpies suggest that Ca_(x)S_(1-x)compounds undergo multiple phase transitions and eventually decompose into elemental Ca and S,challenging the traditional view that pressure stabilizes and densifies compounds.The analysis of formation enthalpy indicates that an increase in pressure leads to a rise in internal energy and the PV term,resulting in thermodynamic instability.Bader charge analysis reveals that this phenomenon is attributed to a decrease in charge transfer under high pressure.The activation of Ca-3d orbitals is significantly enhanced under pressure,leading to competition with Ca-4s orbitals and S-3p orbitals.This may cause the formation enthalpy minimum on the convex hull to shift sequentially from CaS to CaS_(3),then to Ca_(3)S and Ca_(2)S,and finally back to CaS.These findings provide critical insights into the behavior of alkaline-earth metal sulfides under high pressure,with implications for the synthesis and application of novel materials under extreme conditions and for understanding element distribution in planetary interiors.
基金Supported by Talent Scientific Research Start-up Foundation of Wannan Medical College,No.WYRCQD2023045.
文摘BACKGROUND The early diagnosis rate of pancreatic ductal adenocarcinoma(PDAC)is low and the prognosis is poor.It is important to develop an interpretable noninvasive early diagnostic model in clinical practice.AIM To develop an interpretable noninvasive early diagnostic model for PDAC using plasma extracellular vesicle long RNA(EvlRNA).METHODS The diagnostic model was constructed based on plasma EvlRNA data.During the process of establishing the model,EvlRNA-index was introduced,and four algorithms were adopted to calculate EvlRNA-index.After the model was successfully constructed,performance evaluation was conducted.A series of bioinformatics methods were adopted to explore the potential mechanism of EvlRNA-index as the input feature of the model.And the relationship between key characteristics and PDAC were explored at the single-cell level.RESULTS A novel interpretable machine learning framework was developed based on plasma EvlRNA.In this framework,a two-layer classifier was established.A new concept was proposed:EvlRNA-index.Based on EvlRNA-index,a cancer diagnostic model was established,and a good diagnostic effect was achieved.The accuracy of PDACandCPvsHealth-Probabilistic PCA Index-SVM(PDAC and chronic pancreatitis vs health-probabilistic principal component analysis index-support vector machine)(1-18)was 91.51%,with Mathew’s correlation coefficient 0.7760 and area under the curve 0.9560.In the second layer of the model,the accuracy of PDACvsCP-Probabilistic PCA Index-RF(PDAC vs chronic pancreatitis-probabilistic principal component analysis index-random forest)(2-17)was 93.83%,with Mathew’s correlation coefficient 0.8422 and area under the curve 0.9698.Forty-nine PDAC-related genes were identified,among which 16 were known,inferring that the remaining ones were also PDAC-related genes.CONCLUSION An interpretable two-layer machine learning framework was proposed for early diagnosis and prediction of PDAC based on plasma EvlRNA,providing new insights into the clinical value of EvlRNA.
基金supported by the National Key Research and Development Program of China(No.2022YFD1700102)the National Natural Science Foundation of China(No.U20A20108)the Hunan Provincial Natural Science Foundation of China(No.2021JJ30357).
文摘Organic matter(OM)derived from the decomposition of crop residues plays a key role as a sorbent for cadmium(Cd)immobilization.Few studies have explored the straw decomposition processes with the presence of minerals,and the effect of newly generated organomineral complexes on heavy metal adsorption.In this study,we investigated the variations in structure and composition during the rice straw decomposition with or without minerals(goethite and kaolinite),as well as the adsorption behavior and mechanisms by which straw decomposition affects Cd immobilization.The degree of humification of extracted straw organic matter was assessed using excitation-emission matrix(EEM)fluorescence and Ultraviolet-visible spectroscopy(UV-vis),while employing FTIR spectroscopy and XPS to characterize the adsorption mechanisms.The spectra analysis revealed the enrichment of highly aromatic and hydrophobic components,indicating that the degree of straw decomposition and humification were further intensified during incubation.Additionally,the existence of goethite(SG)accelerated the humification of OM.Sorption experiments revealed that the straw humification increased Cd adsorption capacity.Notably,SG exhibited significantly higher adsorption performance compared to the organic matter without minerals(RS)and the existence of kaolinite(SK).Further analysis using FT-IR spectroscopy and XPS verified that the primarymechanisms involved in Cd immobilization were complexion with—OH and—COOH,as well as the formation of Cd-πbinds with aromatic C=C on the surface of solid OMs.These findings will facilitate understanding the interactions of the rice straw decomposing with soil minerals and its remediation effect on Cd-contaminated farmland.
基金supported as part of the"2024 Groundwater Basic Survey Project,"funded by the Ministry of Environment(Project No.20240123-001)。
文摘Earthquakes can cause significant hazards,leading to loss of human life and property damage.In this study,we applied the masking empirical mode decomposition(EMD)to the Gyeongju earthquake(ML 5.8)in South Korea and the Abe Pura earthquake(ML 7.0)in Indonesia and then revealed that the 7th intrinsic mode function(IMF7),composed of waveforms with a period band of approximately 150 to 750 minutes.
基金supported by the Basic Scientific Research Business Fund Project of Higher Education Institutions in Heilongjiang Province(145409601)the First Batch of Experimental Teaching and Teaching Laboratory Construction Research Projects in Heilongjiang Province(SJGZ20240038).
文摘In the wake of major natural disasters or human-made disasters,the communication infrastruc-ture within disaster-stricken areas is frequently dam-aged.Unmanned aerial vehicles(UAVs),thanks to their merits such as rapid deployment and high mobil-ity,are commonly regarded as an ideal option for con-structing temporary communication networks.Con-sidering the limited computing capability and battery power of UAVs,this paper proposes a two-layer UAV cooperative computing offloading strategy for emer-gency disaster relief scenarios.The multi-agent twin delayed deep deterministic policy gradient(MATD3)algorithm integrated with prioritized experience replay(PER)is utilized to jointly optimize the scheduling strategies of UAVs,task offloading ratios,and their mobility,aiming to diminish the energy consumption and delay of the system to the minimum.In order to address the aforementioned non-convex optimiza-tion issue,a Markov decision process(MDP)has been established.The results of simulation experiments demonstrate that,compared with the other four base-line algorithms,the algorithm introduced in this paper exhibits better convergence performance,verifying its feasibility and efficacy.
基金The Australian Research Council(DP200101197,DP230101107).
文摘Formalizing complex processes and phenomena of a real-world problem may require a large number of variables and constraints,resulting in what is termed a large-scale optimization problem.Nowadays,such large-scale optimization problems are solved using computing machines,leading to an enormous computational time being required,which may delay deriving timely solutions.Decomposition methods,which partition a large-scale optimization problem into lower-dimensional subproblems,represent a key approach to addressing time-efficiency issues.There has been significant progress in both applied mathematics and emerging artificial intelligence approaches on this front.This work aims at providing an overview of the decomposition methods from both the mathematics and computer science points of view.We also remark on the state-of-the-art developments and recent applications of the decomposition methods,and discuss the future research and development perspectives.
基金National Key Research and Development Program of China,Grant/Award Number:2023YFB2903904。
文摘Heterogeneous graphs generally refer to graphs with different types of nodes and edges.A common approach for extracting useful information from heterogeneous graphs is to use meta-graphs,which can be seen as a special kind of directed acyclic graph with same node and edge types as the heterogeneous graph.However,how to design proper metagraphs is challenging.Recently,there have been many works on learning suitable metagraphs from a heterogeneous graph.Existing methods generally introduce continuous weights for edges that are independent of each other,which ignores the topological structures of meta-graphs and can be ineffective.To address this issue,the authors propose a new viewpoint from tensor on learning meta-graphs.Such a viewpoint not only helps interpret the limitation of existing works by CANDECOMP/PARAFAC(CP)decomposition,but also inspires us to propose a topology-aware tensor decomposition,called TENSUS,that reflects the structure of DAGs.The proposed topology-aware tensor decomposition is easy to use and simple to implement,and it can be taken as a plug-in part to upgrade many existing works,including node classification and recommendation on heterogeneous graphs.Experimental results on different tasks demonstrate that the proposed method can significantly improve the state-of-the-arts for all these tasks.
