学 术

分享到微信 ×
打开微信“扫一扫”
即可将网页分享至朋友圈
博士后学术沙龙(第14期)
文:唐小青 来源:党委教师工作部、人力资源部(教师发展中心) 时间:2017-06-14 4032

  为搭建我校博士后之间的学术交流平台,促进学术水平提升,学校博士后管理办公室组织开展博士后学术沙龙活动。本次沙龙由我校博士后刘勇、邱曦伟和罗咏颉分享其研究成果,诚挚邀请感兴趣的师生参加。

  一、时 间:2017年6月15日(周四)15:00-16:30

  二、地 点:清水河校区经管楼宾诺咖啡

  三、主办单位:电子科技大学博士后管理办公室

  四、承办单位:计算机科学与工程学院  电子科技大学博士后联谊会

  五、活动安排:

  报告一:

  (1)主题:Application of Big Data Technologies in the Field of Oil and Gas Sales

  (2)主讲人:信息与软件工程学院博士后 刘勇

  (3)交流内容:

  Recently, oilfield enterprises carry out the integration of data resources as the focus of information technology. Exploration, development and sales of oil and gas in China can produce more than 10TB of data per day. As data information is no longer expensive, how to get valuable information from the massive information becomes very expensive. On the one hand, the valuable information exaction can help oil companies to raise efficiency and reduce costs, on the other hand, it can promote the national energy consumption structure optimization, which will become the basic task of energy data applications. Big data mining for oil and gas sales is becoming the focus of non-oil business of oil companies. Big data technologies for the energy industry will bring new vitality, and will be integrated into the daily life of the public.

  (4)主讲人简介:

  刘勇博士获得西南石油大学油气田开发工程博士学位。他目前是电子科技大学信息与软件工程学院的博士后研究员。他的研究兴趣包括能源和医疗行业应用中的大数据和人工智能技术。

  报告二:

  主题:Reliability-Performance-Energy Correlation Modeling and Resource Optimization for network systems

  主讲人:电子工程学院博士后 邱曦伟

  交流内容:

  Energy-efficient computing has recently attracted much attention, where not only performance but also energy consumption are important metrics to be considered for designing rational resource scheduling strategies. Most of existing approaches for achieving energy efficient computing focus on connecting these two metrics, which however is inadequate because another important factor reliability is not considered. To establish a high fidelity reliability-performance-energy correlation model for the network system, the existed correlations need to be taken into account. Meanwhile, in the large-scale network system, a multi-agent system consisted of multiple local agents and a global agent is usually developed for effectively managing infrastructure. Therefore, we also present an autonomic scheduling mechanism for resource optimization, where the local agents can perform dynamic resource manage for reaching optimum in the net profit metric, and the global agent can focus on dispatching user requests to achieve a global optimization effect.

  主讲人简介:

  邱曦伟博士在2016年获得电子科技大学计算机应用技术的博士学位,现在在电子科技大学协同自主计算实验室(CAC)从事博士后研究工作。他的主要研究方向包括云计算、可靠性、建模优化以及效能计算等。

  报告三:

  (1)主题:Analyze complex networks by sampling, coarse graining and learning representations

  (2)主讲人:计算机科学与工程学院博士后 罗咏颉

  (3)交流内容:

  It is a major research field in physics and information science to answer the questions about how to analyze the complex networks and how to solve complex networks related problems, such as clustering, finding communities, visualization, etc.. Big networks(sometimes also termed as big graph) are the model of real world, e.g. the Internet, power grid of a nation, social network, etc.. Many methods and algorithms for analyzing and computing the properties of small networks can not be used for big networks directly, since the enhancement of networks scale makes it difficult to load data to memory and operate the data. This report sketchs the reasearch advances in network sampling, coarse graining, and learning representations of the complex networks, especially about the spectral graph theory and the spectral graph neural network. It induces the idea to improve the spectral graph neural network by random work method, and I am appreciated to discuss the preliminary thought with you.

  主讲人介绍:

  罗咏劼博士于2016年获得电子科技大学信号与信息处理专业工学博士学位,现为电子科技大学计算机学院全职博士后。他的博士后研究课题为复杂网络的抽样、粗糙化与表示学习。此外,他在攻读博士期间的研究方向为稀疏信号处理与压缩感知。


                 电子科技大学博士后管理办公室

                   计算机科学与工程学院

                    2017年6月14日


编辑:林坤  / 审核:林坤  / 发布:林坤

"