HFTCRNet: Hierarchical Fusion Transformer for Interbank Credit Rating and Risk Assessment.

As a prominent application of deep neural networks in financial literature, bank credit ratings play a pivotal role in safeguarding global economic stability and preventing crises. In the contemporary financial system, interconnectivity among banks has reached unprecedented levels. However, many existing credit risk models continue to assess each bank independently, resulting in inevitable suboptimal performance. Thus, developing advanced neural networks to model intricate temporal dynamics and interconnected relationships in the banking system is essential for an effective credit rating and risk assessment learning system. To this end, we propose a novel hierarchical fusion transformer for interbank credit rating and risk assessment (HFTCRNet), which includes the long-term temporal transformer (LT 3 ) module, short-term cross-graph transformer (STCGT) module, attentive risk contagion transformer (ARCT) module, and hierarchical fusion transformer (HFT) module to capture the long-term growth trajectories of banks, the short-term interbank network variance, the potential propagation of risks within interbank network, and integrate these information hierarchically. We further develop an interbank credit rating dataset, encompassing quarterly financial data, interbank lending networks, and key indicators such as credit ratings and systemic risk (SRISK) for 4548 banks from 2016Q1 to 2023Q1. Notably, we also adapt the minimum density algorithm to stabilize the interbank loan network over time, aiding in the analysis of long-term and short-term network effects. Our learning system uses semi-supervised learning to handle labels of varying sparsity, integrating credit ratings and SRISK for a comprehensive assessment of individual bank creditworthiness and systemic interbank risk. Extensive experimental results on our interbank dataset show that HFTCRNet not only outperforms all the baselines in terms of credit rating accuracy but also can evaluate the systemic risk within the interbank network. Code will be available at: https://github.com/AI4Risk/HFTCRNet.

Will the fibers of the future be considered fibers?

This work explores the pivotal breakthroughs and historical developments in fibers over the past century, while also identifying future research directions and emerging trends that promise to shape the future of this field.

Graph Multi-Convolution and Attention Pooling for Graph Classification.

Many studies have achieved excellent performance in analyzing graph-structured data. However, learning graph-level representations for graph classification is still a challenging task. Existing graph classification methods usually pay less attention to the fusion of node features and ignore the effects of different-hop neighborhoods on nodes in the graph convolution process. Moreover, they discard some nodes directly during the graph pooling process, resulting in the loss of graph information. To tackle these issues, we propose a new Graph Multi-Convolution and Attention Pooling based graph classification method (GMCAP). Specifically, the designed Graph Multi-Convolution (GMConv) layer explicitly fuses node features learned from different perspectives. The proposed weight-based aggregation module combines the outputs of all GMConv layers, for adaptively exploiting the information over different-hop neighborhoods to generate informative node representations. Furthermore, the designed Local information and Global Attention based Pooling (LGAPool) utilizes the local information of a graph to select several important nodes and aggregates the information of unselected nodes to the selected ones by a global attention mechanism when reconstructing a pooled graph, thus effectively reducing the loss of graph information. Extensive experiments show that GMCAP outperforms the state-of-the-art methods on graph classification tasks, demonstrating that GMCAP can learn graph-level representations effectively.

Graph Convolutional Networks With Adaptive Neighborhood Awareness.

Graph convolutional networks (GCNs) can quickly and accurately learn graph representations and have shown powerful performance in many graph learning domains. Despite their effectiveness, neighborhood awareness remains essential and challenging for GCNs. Existing methods usually perform neighborhood-aware steps only from the node or hop level, which leads to a lack of capability to learn the neighborhood information of nodes from both global and local perspectives. Moreover, most methods learn the nodes' neighborhood information from a single view, ignoring the importance of multiple views. To address the above issues, we propose a multi-view adaptive neighborhood-aware approach to learn graph representations efficiently. Specifically, we propose three random feature masking variants to perturb some neighbors' information to promote the robustness of graph convolution operators at node-level neighborhood awareness and exploit the attention mechanism to select important neighbors from the hop level adaptively. We also utilize the multi-channel technique and introduce a proposed multi-view loss to perceive neighborhood information from multiple perspectives. Extensive experiments show that our method can better obtain graph representation and has high accuracy.

Observation of the out-of-plane orbital antidamping-like torque.

The out-of-plane antidamping-like orbital torque fosters great hope for high-efficiency spintronic devices. Here we report experimentally the observation of out-of-plane antidamping-like torque that could be generated by z-polarized orbital current in ferromagnetic-metal/oxidized Cu (CuOx) bilayers, which is presented unambiguously by the magnetic field angle dependence of the spin-torque ferromagnetic resonance signal. The CuOx thickness dependence of the orbital torque ratios highlights that the interfacial effect would be responsible for the generation of orbital current. Besides that, the CuOx thickness dependence of the damping parameter further proves the observation of antidamping-like torque. This result contributes to enriching the orbital-related theory of the generation mechanism of the orbital torque.

Hybrid Residual Multiexpert Reinforcement Learning for Spatial Scheduling of High-Density Parking Lots.

Industries, such as manufacturing, are accelerating their embrace of the metaverse to achieve higher productivity, especially in complex industrial scheduling. In view of the growing parking challenges in large cities, high-density vehicle spatial scheduling is one of the potential solutions. Stack-based parking lots utilize parking robots to densely park vehicles in the vertical stacks like container stacking, which greatly reduces the aisle area in the parking lot, but requires complex scheduling algorithms to park and take out the vehicles. The existing high-density parking (HDP) scheduling algorithms are mainly heuristic methods, which only contain simple logic and are difficult to utilize information effectively. We propose a hybrid residual multiexpert (HIRE) reinforcement learning (RL) approach, a method for interactive learning in the digital industrial metaverse, which efficiently solves the HDP batch space scheduling problem. In our proposed framework, each heuristic scheduling method is considered as an expert. The neural network trained by RL assigns the expert strategy according to the current parking lot state. Furthermore, to avoid being limited by heuristic expert performance, the proposed hierarchical network framework also sets up a residual output channel. Experiments show that our proposed algorithm outperforms various advanced heuristic methods and the end-to-end RL method in the number of vehicle maneuvers, and has good robustness to the parking lot size and the estimation accuracy of vehicle exit time. We believe that the proposed HIRE RL method can be effectively and conveniently applied to practical application scenarios, which can be regarded as a key step for RL to enter the practical application stage of the industrial metaverse.

HRegNet: A Hierarchical Network for Efficient and Accurate Outdoor LiDAR Point Cloud Registration.

Point cloud registration is a fundamental problem in 3D computer vision. Outdoor LiDAR point clouds are typically large-scale and complexly distributed, which makes the registration challenging. In this paper, we propose an efficient hierarchical network named HRegNet for large-scale outdoor LiDAR point cloud registration. Instead of using all points in the point clouds, HRegNet performs registration on hierarchically extracted keypoints and descriptors. The overall framework combines the reliable features in deeper layer and the precise position information in shallower layers to achieve robust and precise registration. We present a correspondence network to generate correct and accurate keypoints correspondences. Moreover, bilateral consensus and neighborhood consensus are introduced for keypoints matching, and novel similarity features are designed to incorporate them into the correspondence network, which significantly improves the registration performance. In addition, we design a consistency propagation strategy to effectively incorporate spatial consistency into the registration pipeline. The whole network is also highly efficient since only a small number of keypoints are used for registration. Extensive experiments are conducted on three large-scale outdoor LiDAR point cloud datasets to demonstrate the high accuracy and efficiency of the proposed HRegNet. The source code of the proposed HRegNet is available at https://github.com/ispc-lab/HRegNet2.

Scientization or parametrization? Which system is more advantageous to environmental conservation in the reform of China's state-owned forest farms?

Government leadership and grassroots participation are the most typical institutional arrangements in natural resource management, a topic which has been the subject of vigorous debate for a long time. Individually, these systems are referred to as scientization and parametrization. This paper takes the reform of China's state-owned forest farms (SSFs) as a pointcut, comparing the effects of the 2011 policy (representing scientization) and the 2015 policy (representing parametrization) on environmental conservation. For the period from 2006 to 2018, China's provinces are analyzed via difference-in-differences (DID) and principal components difference-in-differences (PCDID) empirical strategies. The results show that the 2015 policy increased new afforestation by an average of 0.903 units, but the 2011 policy had no significant impact. The influence path of the 2015 policy was to curb corruption, relieve fiscal stress, and stimulate innovation, playing mechanism effects of 20.49%, 14.17%, and 33.55%, respectively. However, the 2015 policy was not ideal in terms of its goal of incentivizing multi-agent participation in investments in conservation. Investors prefer to attempt afforestation projects with shorter payback periods, especially projects related to open forest land. Overall, this study supports the belief that parametric management is a better approach to natural resource management than scientific management, but the latter approach still has limitations. Therefore, we propose to prioritize the promotion of parametric management on the closed forest lands of SSFs, but there is no need to hastily mobilize grassroots participation in open forest land management projects.

A Multichannel Convolutional Decoding Network for Graph Classification.

Graph convolutional networks (GCNs) have shown superior performance on graph classification tasks, and their structure can be considered as an encoder-decoder pair. However, most existing methods lack the comprehensive consideration of global and local in decoding, resulting in the loss of global information or ignoring some local information of large graphs. And the commonly used cross-entropy loss is essentially an encoder-decoder global loss, which cannot supervise the training states of the two local components (encoder and decoder). We propose a multichannel convolutional decoding network (MCCD) to solve the above-mentioned problems. MCCD first adopts a multichannel GCN encoder, which has better generalization than a single-channel GCN encoder since different channels can extract graph information from different perspectives. Then, we propose a novel decoder with a global-to-local learning pattern to decode graph information, and this decoder can better extract global and local information. We also introduce a balanced regularization loss to supervise the training states of the encoder and decoder so that they are sufficiently trained. Experiments on standard datasets demonstrate the effectiveness of our MCCD in terms of accuracy, runtime, and computational complexity.

[Effect of aerobic exercise on AKT/GSK3ß mediated hepatocyte apoptosis in nonalcoholic fatty liver disease].

OBJECTIVE: To investigate the effect of aerobic exercise on AKT/GSK3ß pathway-mediated hepatocyte apoptosis in non-alcoholic fatty liver diseases(NAFLD). METHODS: A total of 30 6-week-old male C57BL/6J mice, and mice were fed adaptively for one week. The control group was fed with ordinary diet, and the model group and model exercise group were fed with high-fat diet until 18 weeks. At the 10th week of the experiment, the model exercise group received aerobic exercise intervention for 8 consecutive weeks until the end of the experiment at the 18th week. Automatic biochemical analyzer to detect serum total cholesterol(TC), triglycerides(TG), alanine aminotransferase(ALT), aspartate aminotransferase(AST), low-density lipoprotein(LDL-C) and high-density lipoprotein(HDL-C) levels. Liver pathological morphology was observed by staining with oil red O and HE. The expression changes of AKT, P-AKT~( Ser473), GSK3ß, P-GSK3ß~(Ser9) and Caspase-3 proteins were detected by western blot, and the apoptosis of hepatocytes was detected by in situ terminal transferase labeling(TUNEL). RESULTS: (1) After intervention, compared with control group, body weight, liver index, serum TC, TG, ALT, AST and LDL-C levels in model group were significantly increased(P<0.01 or P<0.05), while HDL-C level was significantly decreased(P<0.01). Compared with model group, body weight, liver index, serum TC, TG, ALT, AST and LDL-C levels in model exercise group were significantly decreased(P<0.01 or P<0.05), while HDL-C level was significantly increased(P<0.01). (2) Compared with the control group, hepatocyte steatosis and the number of lipid droplets in model group were significantly increased. Compared with the model group, the degree of hepatic adipose degeneration was significantly improved and the number of hepatic lipid droplets was significantly decreased in the model exercise group. (3) Compared with control group, the protein expression levels of P-AKT~(Ser473) and P-GSK3ß~(Ser9) in model group were significantly decreased(P<0.01 or P<0.05), the protein expression levels of Caspase-3 were significantly increased(P<0.05), and the number of hepatocyte apoptosis was significantly increased(P<0.05). Compared with model group, the expression of P-AKT~(Ser473) and P-GSK3ß~(Ser9) protein in model exercise group was significantly increased(P<0.01 or P<0.05), the expression of Caspase-3 protein was significantly decreased(P<0.05), and the number of hepatocyte apoptosis was significantly decreased(P<0.01). CONCLUSION: Aerobic exercise can effectively improve NAFLD, by activating AKT/GSK3ß pathway and increasing the expression of AKT/GSK3ß pathway related molecules, thereby reducing caspase-3 expression and hepatocyte apoptosis.

PSDC: A Prototype-Based Shared-Dummy Classifier Model for Open-Set Domain Adaptation.

Open-set domain adaptation (OSDA) aims to achieve knowledge transfer in the presence of both domain shift and label shift, which assumes that there exist additional unknown target classes not presented in the source domain. To solve the OSDA problem, most existing methods introduce an additional unknown class to the source classifier and represent the unknown target instances as a whole. However, it is unreasonable to treat all unknown target instances as a group since these unknown instances typically consist of distinct categories and distributions. It is challenging to identify all unknown instances with only one additional class. In addition, most existing methods directly introduce marginal distribution alignment to alleviate distribution shift between the source and target domains, failing to learn discriminative class boundaries in the target domain since they ignore categorical discriminative information in the adaptation. To address these problems, in this article, we propose a novel prototype-based shared-dummy classifier (PSDC) model for the OSDA. Specifically, our PSDC introduces an auxiliary dummy classifier to calibrate the source classifier and simultaneously develops a weighted adaptation procedure to align class-wise prototypes for adaptation. We further design a pseudo-unknown learning algorithm to reduce the open-set risk. Extensive experiments on Office-31, Office-Home, and VisDA datasets show that the proposed PSDC can outperform existing methods and achieve the new state-of-the-art performance. The code will be made public.

Sparse-to-Dense Matching Network for Large-Scale LiDAR Point Cloud Registration.

Point cloud registration is a fundamental problem in 3D computer vision. Previous learning-based methods for LiDAR point cloud registration can be categorized into two schemes: dense-to-dense matching methods and sparse-to-sparse matching methods. However, for large-scale outdoor LiDAR point clouds, solving dense point correspondences is time-consuming, whereas sparse keypoint matching easily suffers from keypoint detection error. In this paper, we propose SDMNet, a novel Sparse-to-Dense Matching Network for large-scale outdoor LiDAR point cloud registration. Specifically, SDMNet performs registration in two sequential stages: sparse matching stage and local-dense matching stage. In the sparse matching stage, we sample a set of sparse points from the source point cloud and then match them to the dense target point cloud using a spatial consistency enhanced soft matching network and a robust outlier rejection module. Furthermore, a novel neighborhood matching module is developed to incorporate local neighborhood consensus, significantly improving performance. The local-dense matching stage is followed for fine-grained performance, where dense correspondences are efficiently obtained by performing point matching in local spatial neighborhoods of high-confidence sparse correspondences. Extensive experiments on three large-scale outdoor LiDAR point cloud datasets demonstrate that the proposed SDMNet achieves state-of-the-art performance with high efficiency.

Electrical Manipulation of Orbital Current Via Oxygen Migration in Ni81 Fe19 /CuOx /TaN Heterostructure.

The orbital Hall effect and the interfacial Rashba effect provide new approaches to generate orbital current and spin-orbit torque (SOT) efficiently without the use of heavy metals. However, achieving efficient dynamic control of orbital current and SOT in light metal oxides has proven challenging. In this study, it is demonstrated that a sizable magnetoresistance effect related to orbital current and SOT can be observed in Ni81 Fe19 /CuOx /TaN heterostructures with various CuOx oxidization concentrations. The ionic liquid gating induces the migration of oxygen ions, which modulates the oxygen concentration at the Ni81 Fe19 /CuOx interface, leading to reversible manipulation of the magnetoresistance effect and SOT. The existence of a thick TaN capping layer allows for sophisticated internal oxygen ion reconstruction in the CuOx layer, rather than conventional external ion exchange. These results provide a method for the reversible and dynamic manipulation of the orbital current and SOT generation efficiency, thereby advancing the development of spin-orbitronic devices through ionic engineering.

Enhanced Spin Current in Ni81 Fe19 /Cu-CuOx Bilayer with Top and Sideways Oxidization.

Generation and manipulation of spin current are the cores of spintronic devices, which are intensely pursued. Heavy metals with strong spin-orbit coupling are commonly used for the generation of spin current, but are incompatible with the mass production of devices, and the polarization of spin current is limited to be in-plane. Here, it is shown that the spin current with strong out-of-plane polarization component can be generated and transmitted in Ni81 Fe19 /Cu-CuOx bilayer with sideways and top oxidizations. The charge-to-spin current conversion efficiency can be enhanced through the spin currents consisting of both out-of-plane polarization (σz ) and in-plane polarization (σy ) induced by spin-vorticity coupling. Such a spin current is demonstrated to be closely related to the lateral oxidization gradient and can be controlled by changing the temperatures and times of annealing. The finding here provides a novel degree of freedom to produce and control the spin current in spintronic devices.

CEModule: A Computation Efficient Module for Lightweight Convolutional Neural Networks.

Lightweight convolutional neural networks (CNNs) rely heavily on the design of lightweight convolutional modules (LCMs). For an LCM, lightweight design based on repetitive feature maps (LoR) is currently one of the most effective approaches. An LoR mainly involves an extraction of feature maps from convolutional layers (CE) and feature map regeneration through cheap operations (RO). However, existing LoR approaches carry out lightweight improvements only from the aspect of RO but ignore the problems of poor generalization, low stability, and high computation workload incurred in the CE part. To alleviate these problems, this article introduces the concept of key features from a CNN model interpretation perspective. Subsequently, it presents a novel LCM, namely CEModule, focusing on the CE part. CEModule increases the number of key features to maintain a high level of accuracy in classification. In the meantime, CEModule employs a group convolution strategy to reduce floating-point operations (FLOPs) incurred in the training process. Finally, this article brings forth a dynamic adaptation algorithm ( α -DAM) to enhance the generalization of CEModule-enabled lightweight CNN models, including the developed CENet in dealing with datasets of different scales. Compared with the state-of-the-art results, CEModule reduces FLOPs by up to 54% on CIFAR-10 while maintaining a similar level of accuracy in classification. On ImageNet, CENet increases accuracy by 1.2% following the same FLOPs and training strategies.

[Research progress on mechanism of exercise improving non-alcoholic fatty liver disease through regulating endoplasmic reticulum stress].

Non-alcoholic fatty liver disease (NAFLD) is one of the main diseases of metabolic syndrome. With the increasing popularity of NAFLD in the world, the prevention and therapy of NAFLD are facing great challenges. In recent years, scholars at home and abroad have carried out a large number of studies on NAFLD, but its pathogenesis is still unclear. Endoplasmic reticulum stress (ERS) is caused by the accumulation of unfolded or misfolded proteins. In response to ERS, cells help restore normal endoplasmic reticulum function by initiating a protective mechanism known as the unfolded protein response (UPR). Abnormal accumulation of lipids in hepatocytes, aggravated inflammatory response, increased apoptosis of hepatocytes and insulin resistance (IR) are the main pathogenic factors and characteristics of NAFLD, which are closely related to hepatic ERS. A large number of studies have shown that exercise, as a low-cost treatment, can prevent and improve NAFLD effectively, and its mechanism is related to exercise regulating the level of ERS. This paper reviews the research progress on the mechanism of exercise improving NAFLD from the point of view of ERS. The mechanism of exercise improving NAFLD is related to the regulation of hepatocyte lipid metabolism, alleviation of inflammatory reaction, reduction of hepatocyte apoptosis and improvement of IR through regulating ERS.

Learning Transactional Behavioral Representations for Credit Card Fraud Detection.

Credit card fraud detection is a challenging task since fraudulent actions are hidden in massive legitimate behaviors. This work aims to learn a new representation for each transaction record based on the historical transactions of users in order to capture fraudulent patterns accurately and, thus, automatically detect a fraudulent transaction. We propose a novel model by improving long short-term memory with a time-aware gate that can capture the behavioral changes caused by consecutive transactions of users. A current-historical attention module is designed to build up connections between current and historical transactional behaviors, which enables the model to capture behavioral periodicity. An interaction module is designed to learn comprehensive and rational behavioral representations. To validate the effectiveness of the learned behavioral representations, experiments are conducted on a large real-world transaction dataset provided to us by a financial company in China, as well as a public dataset. Experimental results and the visualization of the learned representations illustrate that our method delivers a clear distinction between legitimate behaviors and fraudulent ones, and achieves better fraud detection performance compared with the state-of-the-art methods.

[Effects of aerobic exercise on CNPY2-PERK pathway in the liver of mice with non-alcoholic fatty liver disease].

Objective: To investigate the effects of aerobic exercise on non-alcoholic fatty liver (NAFLD) induced by high fat and the mechanism of CNPY2-PERK pathway. Methods: Eight-week-old male C57BL/6J mice were randomly divided into four groups: the control group (C), the C+ aerobic exercise group (CE), the model group (M) and the M+ aerobic exercise group (ME). Mice in group C and CE were given normal diet, while mice in group M and ME were given high-fat diet (60 cal % fat). The mice were fed continuously for 18 weeks until the end of the experiment, and the serum and liver samples were collected. Both CE and ME group performed an aerobic treadmill training from the 10th week (12 m/min, 60 min/ day, 5 days/week, for 8 weeks). The serum levels of total cholesterol (TC), triacylglycerol (TG), high-density lipoprotein cholesterol (HDL-c), low-density lipoprotein cholesterol (LDL-c), alanine aminotransferase (ALT), aspartate aminotransferase (AST) were detected. The pathological morphology of the liver was observed. The relative expressions of CNPY2, PERK, p-eIF2a, CHOP, CNPY2 mRNA and PERK mRNA, and the positive expressions of CNPY2 and PERK were measured. Results: Compared with group C, the serum levels of LDL-c, TC, TG, ALT and AST in group M were increased significantly (P＜0.05), while the HDL-c level was decreased significantly (P＜0.05). The liver tissues of mice showed obvious hepatic steatosis, the number of lipid droplets in liver cells was increased, and the expressions of CNPY2, CNPY2mRNA, PERK, PERKmRNA, p-eIF2a, CHOP, and the positive expressions of both CNPY2 and PERK in liver were increased (P＜0.05). However, the above indexes showed no significant difference in CE group (P＞0.05). Compared with group M, the serum levels of LDL-c, TC, TG, ALT and AST in group ME were decreased (P＜0.05). The fatty degeneration of liver tissue and the number of lipid droplets in liver cells in mice was reduced, and the expressions of CNPY2, CNPY2 mRNA, PERK, PERK mRNA, p-eIF2a, CHOP, and the positive expressions of CNPY2 and PERK in liver tissue were decreased (P＜0.05). Conclusion: The CNPY2-PERK pathway is involved in the formation of NAFLD. Aerobic exercise can effectively ameliorate NAFLD, and the mechanisms may be related to the reduction of CNPY2-PERK pathway-related molecule expressions by aerobic exercise.

Transcriptome Profiling to the Effects of Drought Stress on Different Propagation Modes of Tea Plant (Camellia sinensis).

Tea plant (Camellia sinensis) is an important economic beverage crop. Drought stress seriously affects the growth and development of tea plant and the accumulation of metabolites, as well as the production, processing, yield and quality of tea. Therefore, it is necessary to understand the reaction mechanism of tea plant under drought conditions and find efficient control methods. Based on transcriptome sequencing technology, this study studied the difference of metabolic level between sexual and asexual tea plants under drought stress. In this study, there were multiple levels of up-regulation and down-regulation of differential genes related to cell composition, molecular function and biological processes. Transcriptomic data show that the metabolism of tea plants with different propagation modes of QC and ZZ is different under drought conditions. In the expression difference statistics, it can be seen that the differential genes of QC are significantly more than ZZ; GO enrichment analysis also found that although differential genes in biological process are mainly enriched in the three pathways of metabolic, single organism process and cellular process, cellular component is mainly enriched in cell, cell part, membrane, and molecular function, and binding, catalytic activity, and transporter activity; the enrichment order of differential genes in these pathways is different in QC and ZZ. This difference is caused by the way of reproduction. The further study of these differential genes will lay a foundation for the cultivation methods and biotechnology breeding to improve the quality of tea.

Net Learning.

Graph neural networks, which generalize deep learning to graph-structured data, have achieved significant improvements in numerous graph-related tasks. Petri nets (PNs), on the other hand, are mainly used for the modeling and analysis of various event-driven systems from the perspective of prior knowledge, mechanisms, and tasks. Compared with graph data, net data can simulate the dynamic behavioral features of systems and are more suitable for representing real-world problems. However, the problem of large-scale data analysis has been puzzling the PN field for decades, and thus, limited its universal applicability. In this article, a framework of net learning (NL) is proposed. NL contains the advantages of PN modeling and analysis with the advantages of graph learning computation. Then, two kinds of NL algorithms are designed for performance analysis of stochastic PNs, and more specifically, the hidden feature information of the PN is obtained by mapping net information to the low-dimensional feature space. Experiments demonstrate the effectiveness of the proposed model and algorithms on the performance analysis of stochastic PNs.