-
Notifications
You must be signed in to change notification settings - Fork 0
/
seminar.html
3 lines (3 loc) · 15.5 KB
/
seminar.html
1
2
3
<!doctype html><html><head><meta charset="utf-8"><meta http-equiv="X-UA-Compatible" content="IE=edge"><title>SUSTech DBGroup</title><meta name="description" content=""><meta name="viewport" content="width=device-width,initial-scale=1"><meta name="robots" content="all,follow"><link rel="stylesheet" href="vendor/bootstrap/css/bootstrap.min.css"><link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Roboto:300,400,700"><link rel="stylesheet" href="css/pe-icon-7-stroke.css"><link rel="stylesheet" href="vendor/lightbox2/css/lightbox.min.css"><link rel="stylesheet" href="vendor/leaflet/leaflet.css"><link rel="stylesheet" href="css/style.default.css" id="theme-stylesheet"><link rel="stylesheet" href="css/custom.css"><link rel="shortcut icon" href="img/favicon.png"><!--[if lt IE 9]>
<script src="https://oss.maxcdn.com/html5shiv/3.7.3/html5shiv.min.js"></script>
<script src="https://oss.maxcdn.com/respond/1.4.2/respond.min.js"></script><![endif]--><script src="vendor/vue/vue.min.js"></script><meta http-equiv="Expires" content="0"><meta http-equiv="Pragma" content="no-cache"><meta http-equiv="Cache-control" content="no-cache"><meta http-equiv="Cache" content="no-cache"><script defer="defer" src="js/seminar.d85dcafbfea62ba8751a.js"></script></head><body><header class="header"><nav class="navbar navbar-expand-lg py-lg-0"><div class="container"><a href="index.html"><img src="img/dbgroup-logo.png" alt="DBGroup" height="50px"/> </a><button type="button" data-toggle="collapse" data-target="#navbarSupportedContent" aria-controls="navbarSupportedContent" aria-expanded="false" aria-label="Toggle navigation" class="navbar-toggler navbar-toggler-right"><span class="text-uppercase font-weight-bold mr-2">Menu</span><i class="fas fa-bars"></i></button><div id="navbarSupportedContent" class="collapse navbar-collapse"><ul class="navbar-nav ml-auto d-lg-flex align-items-lg-center"><li class="nav-item"><a href="index.html" class="nav-link font-weight-bold text-uppercase px-lg-3 py-lg-4">Home</a></li><li class="nav-item"><a href="seminar.html" class="nav-link font-weight-bold text-uppercase px-lg-3 py-lg-4 active">Seminar</a></li></ul></div></div></nav></header><section class="py-2 bg-cover bg-gray"><div class="container"><div aria-label="breadcrumb" class="nav"><ol class="breadcrumb"><li class="breadcrumb-item"><a href="./">Home</a></li><li aria-current="page" class="breadcrumb-item active">Seminar</li></ol></div><h1 class="">Seminars</h1></div></section><section class="py-5"><div class="container py-2"><div style="width: 100%; display: flex; justify-content: center"><h2 class="section-title mb-4 lined">2022</h2></div><div class="row publication-list"><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/invitation/谢志强.jpeg" alt=""></div></div><div class="col-md-10"><p class="publication-title">Tailoring Abstractions for Emerging Machine Learning Workloads</p><p class="publication-location">2022-05-28</p><p class="publication-author" title="谢志强，斯坦福大学博士生">Speaker: 谢志强，斯坦福大学博士生</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-0')">abstract</a></div><div class="abstract" id="abstract-0-0"><div><p>Performing emerging machine learning computation on hardware accelerators efficiently is challenging. Existing frameworks often adopt several standard intermediate representation layers to bridge the gap between model description and hardware execution units, including high level interfaces, computational graphs and kernel implementation. While this kind of design simplifies the system stack implementation, it can incur fundamental performance limitations. In this talk, I will present Rammer (OSDI ‘20) and Graphiler (GNNSys ‘21), two compiler stacks for ordinary Deep Neural Networks (DNNs) and Graph Neural Networks (GNNs) respectively.</p></div></div></div></div></div><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/member/易岚菁.jpeg" alt=""></div></div><div class="col-md-10"><p class="publication-title">边缘计算场景下的高效时空分位数监测</p><p class="publication-location">2022-05-19</p><p class="publication-author" title="易岚菁, 南科大本科生">Speaker: 易岚菁, 南科大本科生</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-1')">abstract</a></div><div class="abstract" id="abstract-0-1"><div><p>分位数计算 (Quantile Computation) 是了解数据集分布的基本手段，也是诸多数据查询处理、分析挖掘以及机器学习任务的基础模块。当下，越来越多的物理信息 (如温湿度、车速、噪声等) 被传感器所采集，形成了高速高密度的时空大数据。目前，集中式的分位数计算方法难以实现对高速高密度物联网数据的高响应能力 (responsiveness)，而少数的分布式的、针对无线传感网络的方案则未考虑弹性多变的多查询用户场景。本工作提出了一个适应物联网环境的、基于边缘计算 (edge computing) 和数据摘要技术(data sketching) 的处理框架，以高效完成时空分位数监测的查询应答。</p></div></div></div></div></div><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/member/李润中.jpeg" alt=""></div></div><div class="col-md-10"><p class="publication-title">Resource Management and Templates in C++</p><p class="publication-location">2022-04-14</p><p class="publication-author" title="李润中, 南科大-香港理工大学联合培养博士生">Speaker: 李润中, 南科大-香港理工大学联合培养博士生</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-2')">abstract</a></div><div class="abstract" id="abstract-0-2"><div><p>在使用C++时，程序员往往需要对内存进行手动管理，这正是C++有别于其他具有高级特性的编程语言之处。这为C++带来了灵活的资源控制粒度和较低的资源消耗，但同时为程序设计引入了不小的麻烦。本次报告主要通过对C++现代资源管理设施的介绍，展示现代C++的资源管理理念，以及消除不安全资源管理的手段。本次报告同时简要地介绍如何利用C++模板元编程的高级新特性来减少重复代码的编写。</p></div></div></div></div></div><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/member/张家澍.jpg" alt=""></div></div><div class="col-md-10"><p class="publication-title">Plotting in Python</p><p class="publication-location">2022-04-07</p><p class="publication-author" title="张家澍, 香港理工大学博士生">Speaker: 张家澍, 香港理工大学博士生</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-3')">abstract</a></div><div class="abstract" id="abstract-0-3"><div><p>在学术论文中，我们常常需要绘制大量的点图，线图和柱图来呈现实验结果，好的作图方法允许我们在更好地呈现实验结果的同时提高工作的效率。我们有许多方式绘制论文中所需要的点图，线图和柱图，如通过Office软件，MATLAB和Origin。在本次学术论坛中，我们介绍一种跨平台的作图方法，使用Python进行作图。我们结合一些例子，向大家展示如何使用Python进行作图，如何微调图，以及如何将图插入到论文中。</p></div></div></div></div></div><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/member/向隆.jpeg" alt=""></div></div><div class="col-md-10"><p class="publication-title">Introduction to NVM and RDMA</p><p class="publication-location">2022-03-31</p><p class="publication-author" title="向隆, 南方科技大学博士生">Speaker: 向隆, 南方科技大学博士生</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-4')">abstract</a></div><div class="abstract" id="abstract-0-4"><div><p>RDMA，NVM，GPU等新型硬件的出现为提升数据库性能带了新的机遇和挑战，如何结合新硬件的特性设计出高效的数据库系统是值得深入研究的问题。本次报告主要针对RDMA和NVM这两种硬件的特性作以介绍，并介绍上交IPADS实验室关于使用RDMA和NVM加速数据库事务处理的两个工作（DrTM+HOSDI’18,RDpMAATC’21），从而学习如何更好使用RDMA和NVM来加速数据库应用</p></div></div></div></div></div><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/member/刘昊天.jpeg" alt=""></div></div><div class="col-md-10"><p class="publication-title">Performance Isolation in multi-tenant DBaaS</p><p class="publication-location">2022-03-24</p><p class="publication-author" title="刘昊天, 南方科技大学博士生">Speaker: 刘昊天, 南方科技大学博士生</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-5')">abstract</a></div><div class="abstract" id="abstract-0-5"><div><p>在数据库即服务(DBaaS)中，供应商托管资源的管理以及数据的计算。为了节约成本，供应商通常需要增加对多租户的支持，即一台机器的计算、存储等资源被多个租户所共享。在此场景下，如何避免租户之间的资源竞争，满足每个租户SLA的同时实现资源的隔离成为一个重点问题。本次报告主要回顾实现多租户及资源隔离的不同技术以及对其中一个重点工作SQLVM进行详细讨论。</p></div></div></div></div></div><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/member/王方.jpeg" alt=""></div></div><div class="col-md-10"><p class="publication-title">Introduction to Cloud-native Database</p><p class="publication-location">2022-03-17</p><p class="publication-author" title="王方, 香港理工大学博士生">Speaker: 王方, 香港理工大学博士生</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-6')">abstract</a></div><div class="abstract" id="abstract-0-6"><div><p>Cloud-native databases become increasingly important for the era of cloud computing, due to the needs for elasticity and on-demand usage by various applications. This talk would introduce the elegant designs of Aurora from AWS and PolarDB from Alibaba.</p></div></div></div></div></div><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/invitation/张彦春.jpeg" alt=""></div></div><div class="col-md-10"><p class="publication-title">大数据/AI技术主力精神健康/神经系统疾病分析, 诊断</p><p class="publication-location">2022-01-06</p><p class="publication-author" title="张彦春, 广州大学特聘教授、澳大利亚维多利亚大学名誉教授">Speaker: 张彦春, 广州大学特聘教授、澳大利亚维多利亚大学名誉教授</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-7')">abstract</a></div><div class="abstract" id="abstract-0-7"><div><p>医疗健康是目前人工智能和大数据最为关注的领域。人工智能+医疗大数据将对医疗产业赋予新的能量与机会。人工智能技术+医疗大数据是将机器学习和数据挖掘两大核心技术用于医疗健康数据，提高医疗诊治，健康管理水平。体现在智能辅助诊断，疾病风险预测，医学图像分析肿瘤监测，药物挖掘，健康管理等。本报告将从大数据分析/人工智能技术应用的角度出发，对, 脑电信号，心电信号等医学数据进行分析，挖掘，探讨人体和疾病各因素之间的关系。应用场景将包括睡眠健康/精神健康，重症监测预警等</p></div></div></div></div></div><div class="seminar-item"><div class="row seminar-section"><div class="col-md-2 seminar-wrapper mt-2"><div class="seminar-container rounded-circle"><img style="left: 0px; top: 0px; transform: scale(1)" src="img/invitation/王鑫.png" alt=""></div></div><div class="col-md-10"><p class="publication-title">知识图谱存储与查询技术新进展</p><p class="publication-location">2022-01-06</p><p class="publication-author" title="王鑫, 天津大学智能与计算学部教授、博导，人工智能学院副院长。中国计算机学会杰出会员、中国计算机学会信息系统专委会秘书长、数据库专委会执行委员;中国中文信息学会语言与知识计算专委会委员。">Speaker: 王鑫, 天津大学智能与计算学部教授、博导，人工智能学院副院长。中国计算机学会杰出会员、中国计算机学会信息系统专委会秘书长、数据库专委会执行委员;中国中文信息学会语言与知识计算专委会委员。</p><div class="publication-btn-group"><a href="javascript:;" class="btn btn-primary my-2" onclick="toggleAbstractDisplay('abstract-0-8')">abstract</a></div><div class="abstract" id="abstract-0-8"><div><p>知识图谱是人工智能的重要基石。各领域大规模知识图谱的构建与应用己对知识图谱存储与查询技术提出了新挑战。本报告将介绍知识图谱存储与查询技术的最新进展，包括：知识图谱数据模型和查询语言、知识图谱存储管理和查询处理、知识图谱数据库管理系统等。同时介绍团队近年来在知识图谱存储与查询方向上所取得的研究成果。</p></div></div></div></div></div></div></div></section><footer style="background: #eee;" class="footer"><div class="bg-dark py-2"><div class="container"><div class="row"><div class="" style="width: 100%; display: flex; align-items: center;justify-content: space-around;flex-wrap: wrap;"><p class="small text-gray mb-0 mx-3" style="white-space: nowrap;">© 2022 All rights reserved</p><p class="small text-gray mb-0 mx-3" style="white-space: nowrap;">Template by <a href="https://bootstrapious.com" class="text-gray">Bootstrapious.com</a></p></div></div></div></div></footer><script src="vendor/jquery/jquery.min.js"></script><script src="vendor/bootstrap/js/bootstrap.bundle.min.js"></script><script src="vendor/lightbox2/js/lightbox.min.js"></script><script src="vendor/leaflet/leaflet.js"></script><link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.7.1/css/all.css" integrity="sha384-fnmOCqbTlWIlj8LyTjo7mOUStjsKC4pOpQbqyi7RrhN7udi9RwhKkMHpvLbHG9Sr" crossorigin="anonymous"></body></html>