Search

Your search keyword '"Yang, Jingwen"' showing total 534 results
Start Over Author "Yang, Jingwen" Search Limiters Full Text
534 results on '"Yang, Jingwen"'

1. Fast Crystal Tensor Property Prediction: A General O(3)-Equivariant Framework Based on Polar Decomposition

Author

Hua, Haowei, Lin, Wanyu, and Yang, Jingwen

Subjects
Computer Science - Computational Engineering, Finance, and Science
Abstract

Predicting the tensor properties of crystalline materials is a fundamental task in materials science. Unlike single-value property prediction, which is inherently invariant, tensor property prediction requires maintaining $O(3)$ group tensor equivariance. This equivariance constraint often introduces tremendous computational costs, necessitating specialized designs for effective and efficient predictions. To address this limitation, we propose a general $O(3)$-equivariant framework for fast crystal tensor property prediction, called GoeCTP. Our framework is efficient as it does not need to impose equivalence constraints onto the network architecture. Instead, GoeCTP captures the tensor equivariance with a simple external rotation and reflection (R&R) module based on polar decomposition. The crafted external R&R module can rotate and reflect the crystal into an invariant standardized crystal position in space without introducing extra computational cost. We show that GoeCTP is general as it is a plug-and-play module that can be smoothly integrated with any existing single-value property prediction framework for predicting tensor properties. Experimental results indicate that GoeCTP achieves higher prediction performance and runs 13$\times$ faster compared to existing state-of-the-art methods in elastic benchmarking datasets, underscoring its effectiveness and efficiency.

Published
2024

2. Cross-Domain Policy Adaptation by Capturing Representation Mismatch

Author

Lyu, Jiafei, Bai, Chenjia, Yang, Jingwen, Lu, Zongqing, and Li, Xiu

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract

It is vital to learn effective policies that can be transferred to different domains with dynamics discrepancies in reinforcement learning (RL). In this paper, we consider dynamics adaptation settings where there exists dynamics mismatch between the source domain and the target domain, and one can get access to sufficient source domain data, while can only have limited interactions with the target domain. Existing methods address this problem by learning domain classifiers, performing data filtering from a value discrepancy perspective, etc. Instead, we tackle this challenge from a decoupled representation learning perspective. We perform representation learning only in the target domain and measure the representation deviations on the transitions from the source domain, which we show can be a signal of dynamics mismatch. We also show that representation deviation upper bounds performance difference of a given policy in the source domain and target domain, which motivates us to adopt representation deviation as a reward penalty. The produced representations are not involved in either policy or value function, but only serve as a reward penalizer. We conduct extensive experiments on environments with kinematic and morphology mismatch, and the results show that our method exhibits strong performance on many tasks. Our code is publicly available at https://github.com/dmksjfl/PAR., Comment: ICML 2024

Published
2024

3. DVS-RG: Differential Variable Speed Limits Control using Deep Reinforcement Learning with Graph State Representation

Author

Yang, Jingwen, Wang, Ping, Golpayegani, Fatemeh, and Wang, Shen

Subjects
Electrical Engineering and Systems Science - Systems and Control
Abstract

Variable speed limit (VSL) control is an established yet challenging problem to improve freeway traffic mobility and alleviate bottlenecks by customizing speed limits at proper locations based on traffic conditions. Recent advances in deep reinforcement learning (DRL) have shown promising results in solving VSL control problems by interacting with sophisticated environments. However, the modeling of these methods ignores the inherent graph structure of the traffic state which can be a key factor for more efficient VSL control. Graph structure can not only capture the static spatial feature but also the dynamic temporal features of traffic. Therefore, we propose the DVS-RG: DRL-based differential variable speed limit controller with graph state representation. DVS-RG provides distinct speed limits per lane in different locations dynamically. The road network topology and traffic information(e.g., occupancy, speed) are integrated as the state space of DVS-RG so that the spatial features can be learned. The normalization reward which combines efficiency and safety is used to train the VSL controller to avoid excessive inefficiencies or low safety. The results obtained from the simulation study on SUMO show that DRL-RG achieves higher traffic efficiency (the average waiting time reduced to 68.44\%) and improves the safety measures (the number of potential collision reduced by 15.93\% ) compared to state-of-the-art DRL methods.

Published
2024

4. TimeSeriesBench: An Industrial-Grade Benchmark for Time Series Anomaly Detection Models

Author

Si, Haotian, Li, Jianhui, Pei, Changhua, Cui, Hang, Yang, Jingwen, Sun, Yongqian, Zhang, Shenglin, Li, Jingjing, Zhang, Haiming, Han, Jing, Pei, Dan, and Xie, Gaogang

Subjects
Computer Science - Machine Learning
Abstract

Time series anomaly detection (TSAD) has gained significant attention due to its real-world applications to improve the stability of modern software systems. However, there is no effective way to verify whether they can meet the requirements for real-world deployment. Firstly, current algorithms typically train a specific model for each time series. Maintaining such many models is impractical in a large-scale system with tens of thousands of curves. The performance of using merely one unified model to detect anomalies remains unknown. Secondly, most TSAD models are trained on the historical part of a time series and are tested on its future segment. In distributed systems, however, there are frequent system deployments and upgrades, with new, previously unseen time series emerging daily. The performance of testing newly incoming unseen time series on current TSAD algorithms remains unknown. Lastly, the assumptions of the evaluation metrics in existing benchmarks are far from practical demands. To solve the above-mentioned problems, we propose an industrial-grade benchmark TimeSeriesBench. We assess the performance of existing algorithms across more than 168 evaluation settings and provide comprehensive analysis for the future design of anomaly detection algorithms. An industrial dataset is also released along with TimeSeriesBench., Comment: Accepted by ISSRE'24

Published
2024

15. Co-expression of circRNA in oral leukoplakia and oral lichen planus

Author

YANG Jingwen, SONG Yuhan, XU siming, GE Shuyun, and ZHOU Haiwen

Subjects
circrna, circhla-c, oral leukoplakia, oral lichen planus, oral potentially malignant disorders, malignant transformation, biomarkers, high-throughput sequencing, Medicine
Abstract

Objective To find any differentially expressed circRNAs in oral leukoplakia (OLK) and oral lichen planus (OLP), to investigate the possible role of circRNAs in the pathogenesis of these two diseases. Methods This study obtained hospital ethical approval. High-throughput sequencing was used to detect differentially expressed circRNAs in OLK, OLP, oral squamous cell carcinoma and normal oral mucosal tissues. CircRNAs were verified by qRT-PCR, enzyme tolerance assays and Sanger sequencing. GO functional analysis and KEGG pathway analysis were performed to predict the functions of circRNAs in OLP. TargetScan and miRanda were applied to predict targeted miRNAs and mRNAs of circRNAs, and ceRNA networks were mapped. Results A total of 49 circRNAs were differentially expressed in OLK and OLP together, including 30 upregulated and 19 downregulated circRNAs. The five circRNAs confirmed with RT-qPCR, including circHLA-C, circRNF13, circTTN, circSEPN2 and circALDH3A2, were all abnormally expressed in OLK and OLP, among which circHLA-C was a key circRNA with trans splice sites, which was validated by expanding the sample size. ROC curve analysis showed that the area under the circHLA-C curve for predicting OLK was 0.955, and the area under the circHLA-C curve for predicting OLP was 0.988. GO functional analysis showed enrichment of many biological processes related to the immune process. The KEGG pathway with the highest enrichment score was "Natural killer cell mediated cytotoxicity". HLA-C was significantly enriched in these processes/pathways. CeRNA network analysis showed that circHLA-C interacted with a variety of miRNAs, such as hsa-miR-26a-5p, hsa-miR-129-5p, and hsa-miR-29a-3p. Conclusion Many circRNAs were differentially expressed in both OLK and OLP, circHLA-C being the most elevated. CircHLA-C is valuable for the early diagnosis of OLK and OLP and may serve as a potential biomarker for the diagnosis and prognosis of OLK and OLP.

Published
2023
Full Text
View/download PDF

16. Edge-Enhanced Dual Discriminator Generative Adversarial Network for Fast MRI with Parallel Imaging Using Multi-view Information

Author

Huang, Jiahao, Ding, Weiping, Lv, Jun, Yang, Jingwen, Dong, Hao, Del Ser, Javier, Xia, Jun, Ren, Tiaojuan, Wong, Stephen, and Yang, Guang

Subjects
Electrical Engineering and Systems Science - Image and Video Processing, Computer Science - Artificial Intelligence, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

In clinical medicine, magnetic resonance imaging (MRI) is one of the most important tools for diagnosis, triage, prognosis, and treatment planning. However, MRI suffers from an inherent slow data acquisition process because data is collected sequentially in k-space. In recent years, most MRI reconstruction methods proposed in the literature focus on holistic image reconstruction rather than enhancing the edge information. This work steps aside this general trend by elaborating on the enhancement of edge information. Specifically, we introduce a novel parallel imaging coupled dual discriminator generative adversarial network (PIDD-GAN) for fast multi-channel MRI reconstruction by incorporating multi-view information. The dual discriminator design aims to improve the edge information in MRI reconstruction. One discriminator is used for holistic image reconstruction, whereas the other one is responsible for enhancing edge information. An improved U-Net with local and global residual learning is proposed for the generator. Frequency channel attention blocks (FCA Blocks) are embedded in the generator for incorporating attention mechanisms. Content loss is introduced to train the generator for better reconstruction quality. We performed comprehensive experiments on Calgary-Campinas public brain MR dataset and compared our method with state-of-the-art MRI reconstruction methods. Ablation studies of residual learning were conducted on the MICCAI13 dataset to validate the proposed modules. Results show that our PIDD-GAN provides high-quality reconstructed MR images, with well-preserved edge information. The time of single-image reconstruction is below 5ms, which meets the demand of faster processing., Comment: 33 pages, 13 figures, Applied Intelligence

Published
2021

25. A BMP-controlled metabolic/epigenetic signaling cascade directs midfacial morphogenesis

Author

Yang, Jingwen, Zhu, Lingxin, Pan, Haichun, Ueharu, Hiroki, Toda, Masako, Yang, Qian, Hallett, Shawn A., Olson, Lorin E., and Mishina, Yuji

Subjects
Tumor proteins -- Health aspects, Embryonic development -- Health aspects, Mortality, Birth defects -- Health aspects, Stem cells -- Health aspects, Platelet-derived growth factor -- Health aspects, Epigenetic inheritance -- Health aspects, Health care industry
Abstract

Craniofacial anomalies, especially midline facial defects, are among the most common birth defects in patients and are associated with increased mortality or require lifelong treatment. During mammalian embryogenesis, specific instructions arising at genetic, signaling, and metabolic levels are important for stem cell behaviors and fate determination, but how these functionally relevant mechanisms are coordinated to regulate craniofacial morphogenesis remain unknown. Here, we report that bone morphogenetic protein (BMP) signaling in cranial neural crest cells (CNCCs) is critical for glycolytic lactate production and subsequent epigenetic histone lactylation, thereby dictating craniofacial morphogenesis. Elevated BMP signaling in CNCCs through constitutively activated ACVR1 (ca-ACVR1) suppressed glycolytic activity and blocked lactate production via a p53- dependent process that resulted in severe midline facial defects. By modulating epigenetic remodeling, BMP signaling- dependent lactate generation drove histone lactylation levels to alter essential genes of Pdgfra, thus regulating CNCC behavior in vitro as well as in vivo. These findings define an axis wherein BMP signaling controls a metabolic/epigenetic cascade to direct craniofacial morphogenesis, thus providing a conceptual framework for understanding the interaction between genetic and metabolic cues operative during embryonic development. These findings indicate potential preventive strategies of congenital craniofacial birth defects via modulating metabolic-driven histone lactylation., Introduction Craniofacial anomalies, especially midline facial defects, are among the most common birth defects found in patient populations whose appearance are either associated with increased mortality or require lifelong treatment [...]

Published
2024
Full Text
View/download PDF

33. Research progress on m6A-modified circRNA in oral diseases

Author

YANG Jingwen and ZHOU Haiwen

Subjects
circular rna ，, noncoding rnas ，, methylated modification ，, n6-methyladenosine ，, oral squamous cell carcinoma，, oral potential malignant diseases ，, oral ，, tumour ，, biomarkers，, Medicine
Abstract

N6-methyladenosine (m6A) modification is the most abundant RNA modification in eukaryotes and is widely involved in the regulation of RNA nuclear export, splicing, translation and degradation. Increasing evidence shows that m6A methylation modification of circular RNA (circRNA) has great potential in metabolism, immunity and benign and malignant diseases. Here, we review the research status of circRNA m6A methylation modification in physiological processes, malignant tumors and oral diseases. It has been shown that m6A methylation can regulate circRNA by regulating circRNA translation, promoting circRNA nuclear output, and promoting circRNA degradation; M6A-modified circRNA plays a regulatory role in immunity, the reproductive system, myogenesis and development, and malignant tumors; and the M6A methyltransferase METTL3 promotes the occurrence of oral squamous cell carcinoma (OSCC) through m6A modification mediated by YTHDF1. The high expression of METTL14 can effectively reduce the proliferation, migration and invasion of OSCC. There is little research on circRNA m6A methylation modification in the oral cavity, which is limited to preliminary research in ameloblastoma and oral implant bone absorption. It has broad research prospects in the treatment of major oral diseases (such as potential oral malignant diseases and OSCC).

Published
2023
Full Text
View/download PDF

46. Earnings management in European private firms versus public firms

Author

Yang, Jingwen, Jaafar, Aziz, and Hodgkinson, Lynn

Subjects
658.15
Abstract

This thesis presents three papers which empirically investigate earnings management (EM) in European private and public firms. Highlighting the importance of private firms, according to the Federation of European Accountants (2016), there are 99% of firms which are registered as private firms, and are considered as the backbone of the European economy. Similarly, there are around 29 million private firms in the U.S., which account for 86% of U.S. firms and generate almost one-half of the nation’s GDP. To date there is limited evidence about private firms’ EM activities. The thesis attempts to fill this gap by not only analysing EM in private firms, but also by comparing the EM of private firms with that of public firms, and in so doing, assesses whether determinants of EM are similar for both types of firms. The sample used in this study comprises both private and public firms domiciled in 11 European countries for which consolidated financial statements are available for the period from 2001 to 2013. These countries are Belgium, Finland, France, Germany, Greece, Italy, The Netherlands, Portugal, Spain, Sweden, and the U.K. In developing the testable hypotheses, the study employs several theoretical frameworks related to EM including the ‘opportunistic behaviour’ and ‘demand’ theories. The main aim of the first paper is to assess the extent of EM for private and public firms domiciled in Europe. In particular, this paper examines the degree of EM for both private and public firms by considering, in part, the influence of country-level factors including legal enforcement, investor protection and tax systems, as well as the adoption of IFRS. The empirical results show that private firms are more likely to engage in EM than public firms, and the effects of IFRS in lowering EM is more pronounced for public firms than for private firms. Furthermore, country-level factors including legal enforcement, investor protection rules and tax rates are important determinants of EM. Given that enhancing country-level legal enforcement systems and investor protection may reduce the use of EM further in both private and public firms, the current study suggests that standard-setters should consider some of the country-level variables to better guarantee the quality of accounting information across countries. The second paper examines the accrual based EM of private and public firms domiciled in Europe, and questions whether IFRS adoption has improved EM. The paper further examines whether EM between private and public firms differs in the following three scenarios (i) due to the need to meet earnings benchmarks, (ii) when obtaining external financing and (iii) when VI the firm employs a Big4 auditor. The results suggest public firms, in general, have lower EM and report more conservatively than private firms. This result is consistent with the argument that on average public firms need to provide higher quality financial information to investors and other stakeholders relative to private firms. However, the finding suggest public firms’ lower EM is mitigated or decreased in the three settings mentioned above. Finally, this paper observes that the EM of public firms has decreased post-IFRS adoption, supporting the view that IFRS has improved the financial reporting quality of public firms. The third paper assesses the degree of real earnings management (REM) for private and public firms domiciled in Europe. This paper further examines some settings that present clear incentives to manage earnings such as (i) to meet earnings benchmarks, (ii) when obtaining external financing, (iii) when leverage is high, and (iv) when losses are large. The empirical results show that public firms engage in more earnings management through real operating activities, and that REM in public firms increased following IFRS adoption, relative to private firms. Furthermore, this study finds that public firms, overall, face stronger incentives to manage earnings than do private firms’ managers in the four settings mentioned above. The study contributes not only to the EM literature on non-accrual earnings management, but also the current debate on IFRS adoptions.

Published
2018

48. CircCFL1 Promotes TNBC Stemness and Immunoescape via Deacetylation‐Mediated c‐Myc Deubiquitylation to Facilitate Mutant TP53 Transcription.

Author

Wang, Zekun, Li, Yaming, Yang, Jingwen, Sun, Yuhan, He, Yinqiao, Wang, Yuping, Liang, Yiran, Chen, Xi, Chen, Tong, Han, Dianwen, Zhang, Ning, Chen, Bing, Zhao, Wenjing, Wang, Lijuan, Luo, Dan, and Yang, Qifeng

Subjects
CIRCULAR RNA, BREAST cancer, TRIPLE-negative breast cancer, T cells, YAP signaling proteins
Abstract

Triple‐negative breast cancer (TNBC) is the most malignant subtype of breast cancer. TP53, which has a mutation rate of ≈70%–80% in TNBC patients, plays oncogenic roles when mutated. However, whether circRNAs can exert their effects on TNBC through regulating mutant TP53 has not been well evaluated. In this study, circCFL1, which is highly expressed in TNBC cells and tissues and has prognostic potential is identified. Functionally, circCFL1 promoted the proliferation, metastasis and stemness of TNBC cells. Mechanistically, circCFL1 acted as a scaffold to enhance the interaction between HDAC1 and c‐Myc, further promoting the stability of c‐Myc via deacetylation‐mediated inhibition of K48‐linked ubiquitylation. Stably expressed c‐Myc further enhanced the expression of mutp53 in TNBC cells with TP53 mutations by directly binding to the promoter of TP53, which promoted the stemness of TNBC cells via activation of the p‐AKT/WIP/YAP/TAZ pathway. Moreover, circCFL1 can facilitate the immune escape of TNBC cells by promoting the expression of PD‐L1 and suppressing the antitumor immunity of CD8+ T cells. In conclusion, the results revealed that circCFL1 plays an oncogenic role by promoting the HDAC1/c‐Myc/mutp53 axis, which can serve as a potential diagnostic biomarker and therapeutic target for TNBC patients with TP53 mutations. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

50. Heat and Fraud: Evaluating How Room Temperature Influences Fraud Likelihood

Author

Liu, Huanxu, Yang, Jingwen, and Yamada, Yuki

Abstract

Despite the considerable amount of research devoted to understanding fraud, few studies have examined how the physical environment can influence the likelihood of committing fraud. One recent study found a link between room brightness and occurrence of human fraud behaviors. Therefore, the present study aims to investigate how temperature may affect fraud. Based on a power analysis using the effect size observed in a pilot study, we recruited 105 participants and randomly divided them into three temperature groups (warm, medium, and cool). We then counted fraud behaviors in each group and tested for potential significant differences with a Kruskal-Wallis test. Additionally, we used a correlation analysis to determine whether the perceived temperature affected fraud. As a result, regardless of participants' subjective sensory experience or their physical environment, we did not find that temperature-related factors influence the incidence of fraud. We discussed the potential reason for the results and suggested directions for future research.

Published
2020
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results