556 research outputs found
Object-oriented stream programming using Aspects: a high-productivity programming paradigm for hybrid platforms
No full textThe move to massively parallel hybrid platforms, such as multicore CPUs accelerated with heterogeneous GPU co-processing systems, is significantly impacting software programmers because existing programs have to be properly parallelized before they can take advantage of these advanced processing architectures. However, using current programming frameworks such as CUDA leads to tangled source code that combines code for the core computation with that for device and computational kernel management, data transfers between memory spaces, and various optimizations. In this research, we propose a programming system based on the principles of Aspect-Oriented Programming, to un-clutter the code and to improve programmability of these heterogeneous parallel systems. Specifically, we use a standard Object-Oriented language to describe the core computations and aspects to encapsulate all other support functions, such as parallelization granularity and memory access optimization. An aspect-weaving compiler is then used to combine the core OO program with these aspects to generate parallelized programs. This approach modularizes concerns that are hard to manage using conventional programming frameworks such as CUDA, has a small impact on existing program structure as well as performance, and as a result, simplifies the programming of accelerator-based heterogeneous parallel systems. Studies on example programs suggest that programs written using this system can be successfully translated to CUDA programs for execution on a CPU + GPU co-processing system with comparable performance. The performance of the translated code achieved ~80% of the hand-coded CUDA programs. We also introduce a performance model based on Bulk Synchronous Parallel (BSP) to help with quick identification of performance bottlenecks and tuning programs for better performance. This model defines a machine parameter (Machine Characteristic Ratio) and an application parameter (Application Characteristic Ratio) to identify the principle factors that can be used to bound application performance for the hierarchical parallel execution in the GPU co-processing device.Ph.D.Includes bibliographical referencesIncludes vitaby Mingliang Wan
Supplemental Material, Supplementary_tables_and_figures - Prediction of Target Genes and Pathways Associated With Cetuximab Insensitivity in Colorectal Cancer
No full textSupplemental Material, Supplementary_tables_and_figures for Prediction of Target Genes and Pathways Associated With Cetuximab Insensitivity in Colorectal Cancer by Chaoran Yu, Hiju Hong, Jiaoyang Lu, Xuan Zhao, Wenjun Hu, Sen Zhang, Yaping Zong, Zhihai Mao, Jianwen Li, Mingliang Wang, Bo Feng, Jing Sun, and Minhua Zheng in Technology in Cancer Research & Treatment</p
LayerCFL: an efficient federated learning with layer-wised clustering
No full textAbstract
Federated Learning (FL) suffers from the Non-IID problem in practice, which poses a challenge for efficient and accurate model training. To address this challenge, prior research has introduced clustered FL (CFL), which involves clustering clients and training them separately. Despite its potential benefits, CFL can be computationally and communicationally expensive when the data distribution is unknown beforehand. This is because CFL involves the entire neural networks of involved clients in computing the clusters during training, which can become increasingly time-consuming with large-sized models. To tackle this issue, this paper proposes an efficient CFL approach called LayerCFL that employs a Layer-wised clustering technique. In LayerCFL, clients are clustered based on a limited number of layers of neural networks that are pre-selected using statistical and experimental methods. Our experimental results demonstrate the effectiveness of LayerCFL in mitigating the impact of Non-IID data, improving the accuracy of clustering, and enhancing computational efficiency
Identifying marine food web homogenization patterns
No full textEcosystems become increasingly similar to each other, based on species composition. Despite the inevitability of homogenized ecosystems due to global change, few studies have specifically addressed the identification of homogeneous systems in food webs. This study focuses on identifying different patterns of marine food web homogenization by selecting 41 marine food webs and establishing an indicator system. The research classifies the food webs into seven main types based on three different homogenization processes (I, II, III, IV, V, VI, and VII), with approximately 60.1%, 46.3%, and 61% of the homogenization being structural, functional, and resource homogenization, respectively. It highlights the importance of homogenization processes in marine ecosystems, which are mainly driven by interactions between structural and resource homogenization. The research found that Type V exhibited universality in both temporal and spatial dimensions, while Type III also showed universality when the food webs were dominated by resource homogenization. On the other hand, Type I, which was associated with human activities, showed locality when the food web only manifested structural homogenization. Functional homogenization often occurred alongside structural homogenization, as seen in Type IV and Type VII. Yet, when the food web exhibited functional homogenization (Type II), it was directly linked to human activities over the past 20 years. The research aimed to improve the methodology in terms of (a) identifying different food web homogenization patterns; (b) establishing indicators system to quantify food web homogenization; and (c) clarifying the ecological significance of food web homogenization. The study provided a comprehensive understanding of food web homogenization and its associated risks, which could inform nature-based ecosystem management strategies to mitigate the impacts of future climate change
Size and shape-based separation using deterministic lateral displacement microfludic systems
No full textContinuous separation of particles of different sizes and shapes is important in both clinical diagnostics and industrial applications and a number of methods have been developed for such separations. In microfluidic systems, deterministic lateral displacement has proved its great potential in achieving the goal of high throughput and efficient separation. Although it was originally based on transporting the suspension in a convective flow, particles can be also driven with external force fields, thus force-driven DLD (f-DLD) devices were demonstrated. This thesis demonstrates the separation of suspended particles by shape and size using scaled-up macroscopic f-DLD devices, using gravity force and a centrifuge, respectively. In the first set of experiments and for the first time, we demonstrate the potential of gravity-driven DLD devices for the separation of particles of different shapes. Our results show that each type of particle moves in different directions within the array of obstacles in DLD systems, depending on the forcing direction. Interestingly, we show that the migration of the particles can be predicted by the diameter of the inscribed sphere, independent of shape. In the second set of experiments and also for the first time, we combined DLD devices with centrifugal force as the driving field. We show that spherical particles of different sizes are driven to different outlets. We show that at some specific angles this setup provides high separation resolution, but the resolution decreases as the concentration of particles increases.M.S.Includes bibliographical referencesby Mingliang JiangThis work was partially supported by the National Science Foundation Grant no. CBET-1339087
Performance evaluation of urea injection on the emission reduction of dioxins and furans in a commercial municipal solid waste incinerator
No full textPolychlorinated dibenzo-p-furans and dioxins (PCDD/Fs) that pose a great threat to human health are commonly found during the incineration of municipal solid waste. In this study, industrial urea was injected into a commercial MSW incinerator flue gas to evaluate the suppression performances of PCDD/Fs, and the possible inhibition mechanisms were proposed. The results show that the use of urea dramatically reduced the PCDD/Fs emission concentration from 8.87 to 0.63 ng/Nm3, along with a significant decrease in the I-TEQ value (0.26 → 0.047 ng I-TEQ/Nm3), below the Chinese national standard of 0.1 ng I-TEQ/Nm3. The emission reduction cost of the industrial urea was over 41 % lower than that of using activated carbon. Urea molecule poisoned the metal ions, thus seriously inhibiting the de novo synthesis of PCDD/Fs. Furthermore, the decomposition products of urea reduced the concentration of HCl in the flue gas, thereby reducing the formation of Cl2 and hindering the chlorination reaction, which was identified from the reduction of chlorination degree and lower distribution of high-chlorinated PCDD/Fs. This research provided some practice basis and experience to reduce the emission of PCDD/Fs from municipal solid waste incineration for future commercial promotion and application
Understanding diabetic retinopathy
No full textA recent survey shows that diabetes affect 92.4 million people in mainland China [1], among which 16.9 million have diabetic retinopathy (DR) [2]. DR is one of the major causes of blindness in the working age population in both developing and developed countries. Although the exact mechanism by which diabetes causes retinopathy remains elusive, several hypotheses have been postulated to explain pathogenies of the disease. Here we invited five groups of scientists and clinicians to share their recent discoveries and expert opinions in molecular biology, pathophysiology, genetics, and epigenetics of DR.SCI(E)PubMed中国科学引文数据库(CSCD)EDITORIAL [email protected]
Legislating and litigating same sex marriage in China
No full textThis chapter presents research on the strategies that Chinese LGB activists adopt in order to cope with the legal and political environment of the Chinese regime. Anchored in the peculiarities of the Chinese legal environment, the chapter considers two qualitative case studies. The first case involves Li Yinhe, a former professor with the Chinese Academy of Social Sciences, who has repeatedly asked representatives at the National People’s Congress (NPC) to present a Chinese Same-Sex Marriage Bill as an amendment to the current marriage law. The second case comes from the study of the impact of litigation brought by Sun Wenlin and his same sex partner Hu Mingliang against a local civil affairs bureau in 2016. The chapter attends to the legal and rhetorical strategies adopted by these activists in order to carefully frame their objections to ensure they complied with the party-state’s emphasis on marital harmony (hexie) and social stability
MicroRNA-181a Regulates the Proliferation and Differentiation of Hu Sheep Skeletal Muscle Satellite Cells and Targets the YAP1 Gene
No full textMicroRNA (miRNA) is of great importance to muscle growth and development, including
the regulation of the proliferation and differentiation of skeletal muscle satellite cells (SMSCs). In our
research group’s previous study, we found that miR-181a is differentially expressed in the longissimus
dorsi muscle of Hu sheep at different stages. We speculated that miR-181a may participate in the
growth and development process of Hu sheep. To understand the mechanism of miR-181a regulating
the growth and development of Hu sheep skeletal muscle, we extracted skeletal muscle satellite
cells from the longissimus dorsi muscle of 3-month-old Hu sheep fetuses and performed a series of
experiments. Our results showed that miR-181a suppressed SMSCs’ proliferation using QRT-PCR,
Western blot, CCK-8, EDU, and Flow cytometry cycle tests. In addition, QRT-PCR, Western blot,
and immunofluorescence indicated that miR-181a facilitated the differentiation of SMSCs. Then, we
used dual-luciferase reporter gene detection, QRT-PCR, and Western blot to find that the Yes1-related
transcription regulator (YAP1) is the target gene of miR-181a. Our study supplies a research basis for
understanding the regulation mechanism of miR-181a on the growth of Hu sheep skeletal muscle
- …
