A5101914465- Machine Learning in Bioinformatics
- Bioinformatics and Genomic Networks
- Additive Manufacturing Materials and Processes
- Machine Learning in Materials Science
- Data Management and Algorithms
- Algorithms and Data Compression
- Genomics and Phylogenetic Studies
- High Entropy Alloys Studies
- Protein Structure and Dynamics
- High-Temperature Coating Behaviors
- Advanced Database Systems and Queries
- Web Data Mining and Analysis
- Data Mining Algorithms and Applications
- Advanced Computational Techniques and Applications
- Laser-induced spectroscopy and plasma
- Laser-Plasma Interactions and Diagnostics
- Data Quality and Management
- Protein Hydrolysis and Bioactive Peptides
- Ferroelectric and Negative Capacitance Devices
- Magnetic Properties and Synthesis of Ferrites
- Titanium Alloys Microstructure and Properties
- Stochastic Gradient Optimization Techniques
- Advanced Computing and Algorithms
- Error Correcting Code Techniques
- Gene expression and cancer classification
Shanghai University of Engineering Science
2013-2024
China Agricultural University
2024
Shanghai University
2014-2019
Fudan University
2009-2010
Data driven material research is a hot topic in the cross field of artificial intelligence and materials science. The core new prediction to find relationship between structure properties. In this research, machine learning will have important advantages play an role for data. paper,we put forward framework combining feature engineering linear regression correlation properties from High temperature superconductor double perovskites solar cells were employed test feasibility method. former,...
Backgrounds Recent explosion of biological data brings a great challenge for the traditional clustering algorithms. With increasing scale sets, much larger memory and longer runtime are required cluster identification problems. The affinity propagation algorithm outperforms many other classical algorithms is widely applied into researches. However, time space complexity become bottleneck when handling large-scale sets. Moreover, similarity matrix, whose constructing procedure takes long...
Backgrounds Despite continuing progress in X-ray crystallography and high-field NMR spectroscopy for determination of three-dimensional protein structures, the number unsolved newly discovered sequences grows much faster than that determined structures. Protein modeling methods can possibly bridge this huge sequence-structure gap with development computational science. A grand challenging problem is to predict structure from its primary (residues sequence) alone. However, predicting residue...
Exploring the rules of high entropy alloys (HEAs) phase formation has clear guiding significance for design new alloys. An end-to-end framework was proposed to select feature subset and machine learning (ML) model from pool pool, respectively. In this framework, each in is determine its materials based on importance. The final confirmed through evaluation fitting result every subset. This method extracts important factors affecting HEAs. results show that chosen could classify 430 HEAs into...
Abstract An ion beam has the unique feature of being able to deposit its main energy inside a human body kill cancer cells or material. However, conventional accelerators tend be huge in size and cost. In this paper, future intense-laser accelerator is discussed make laser-based compact controllable. The issues laser include efficiency from ions, collimation, spectrum control, bunching, particle control. study, each component designed control quality by simulations. ions improved using solid...
PDF HTML阅读 XML下载 导出引用 引用提醒 基于概率数据流的有效聚类算法 DOI: 作者: 作者单位: 作者简介: 通讯作者: 中图分类号: 基金项目: Supported by the National Basic Research Program of China under Grant No.2005CB321905 (国家重点基础研究发展计划(973)) Effective Clustering Algorithm for Probabilistic Data Stream Author: Affiliation: Fund Project: 摘要 | 图/表 访问统计 参考文献 相似文献 引证文献 资源附件 文章评论...
现有的很多序列聚类算法是基于“局部特征可以表征整个序列”的假设来进行的,即不区分实际应用中序列的整体相似性和局部相似性.这对存在保守子模式的序列,如DNA和蛋白质序列是适用的,但对一些注重整体序列相似性的应用领域,如:在交易数据库中用户购买行为的比较,时间序列数据中全局模式的匹配等,由于难以产生频繁子模式,用基于全局相似性的度量方法进行聚类显得更为合理.此外,在基于局部相似性的序列聚类算法中,选取的局部子模式表征序列的能力也有待进一步提高.由此,针对不同应用领域,分别提出基于整体相似性的序列聚类算法GSClu和基于局部相似性的序列聚类算法LSClu.GSClu和LSClu分别利用带剪枝策略的二分k均值算法和基于有gap约束的强区分度子模式方法对各自领域的序列数据进行聚类.实验采用交易序列数据和蛋白质序列数据,实验结果表明,GSClu和LSClu对各自领域的序列数据具有较快的处理速度和良好的聚类质量.;Many current sequence clustering algorithms are based on the hypothesis that can be...
Protein secondary structure prediction is one of the most important and challenging problems in structural bioinformatics, which has been an essential task determining function proteins. Despite significant progress made recent years, protein maintains its status as prime unsolved computational biology. A novel probability revise table based hidden Markov model (PRT-HMM) method presented this paper with considering dependencies among state transitions. We initial predicted through looking up...
The desktop virtualization environment could propose a solution for the high-definition video processing, which used multi-GPU collaboration and parallel computing. encoding computing is often implemented by multi-threads mode. Based on analysis of GPU multi-level storage structure, data transmission between CPU GPU, pinned memory (zero-copy memory) shared approaches are to shorten time encoded devices. In this paper latest support low-latency remote display's H.264 compression standard...
Network alignment has made great progress and provides new insights into protein function. To visualize the results effectively is difficult because of large scale network. In this paper, a simple tool named Compare (NetCompare) proposed to analyze network results. Firstly, multi-hierarchical clustering method used cluster aligned networks. Secondly, at each level, users can select layout which they prefer display NetCompare offers both global local views. The views briefly regard one subnet...
In inertial fusion, a target DT fuel should be compressed to typically 1000 times the solid density. The implosion nonuniformity is introduced by driver beam's illumination nonuniformity, for example. robust against nonuniformities. this paper, requirement uniformity first discussed. non-uniformity less than few percent. dynamics also briefly reviewed in heavy ion fusion (HIF). Heavy ions deposit their energy inside absorber, and deposition layer rather thick, depending on particle energy....
DNA sequences have the fundamental information for each species and a comparison between is one of basic biological problems. There are large number algorithms applying in sequence alignment. Some approximate alignment, others exact which also plays an important role biology research. In this paper, new parallel algorithm based on FED alignment with MPI proposed. The experimental results indicate that proposed can report matched positions specific improve matching speed MPI, as well reduce...
PDF HTML阅读 XML下载 导出引用 引用提醒 基于参考集索引的高效序列相似性查找算法 DOI: 作者: 作者单位: 作者简介: 通讯作者: 中图分类号: 基金项目: Supported by the National Natural Science Foundation of China under Grant No.0573093 (国家自然科学基金); High-Tech Research and Development Plan No.2006AA02Z329 (国家高技术研究发展计划(863)) Efficient Algorithm for Sequence Similarity Search Based on Reference Indexing Author: Affiliation: Fund Project: 摘要 | 图/表 访问统计 参考文献 相似文献 引证文献 资源附件 文章评论...
Abstract With the development of modern computer technology, data which needs to be stored and processed by application systems becomes more varied. While traditional relational databases still dominate, NoSQL are developing rapidly occupy a place in entire industry. Therefore, applications need connect multiple database at same time. However, different characteristics supported differences query syntax rules make multi-database very complicated. This paper talks about unified middleware...