Annual Report 2017-18

Scholars from the Faculty of Science have demonstrated a world-first breakthrough in the design and synthesis of a smart, globular macromolecular machine vehicle for actively controlled cancer drug delivery. The team comprised scholars in Chemistry and Physics: Associate Professor Dr. Ken Leung, Founding Kwok Yat Wai Endowed Chair of Environmental and Biological Analysis Professor Cai Zongwei, and PhD student Kwan Chak Shing of the Department of Chemistry; Chair Professor Michel A. Van Hove, and Postdoctoral Fellow Dr. Zhao Rundong of the Department of Physics. 理學院的學者設計並合成出一種世界首創的智能球狀分子 機器載體，能控制藥物的釋放及份量，大幅提升標靶藥物 的藥效。研究團隊的成員包括化學系副教授梁湛輝博士、 郭一葦環境與生物分析講座教授蔡宗葦教授和博士生關擇 誠，以及物理系 Michel A. Van Hove 講座教授和博士後研究員 趙潤東博士。 By analysing African lake sedimentary cores, Professor Richard Bernhart Owen of the Department of Geography has discovered strong evidence for links between Africa’s drying climate and human evolution. For example, ancestral human learned new ways of living and developed technologies to cope with climate change. The research was published in one of the world’s most-cited multidisciplinary scientific journals, Proceedings of the National Academy of Sciences of the United States of America . 地理系歐文彬教授分析非洲湖泊的湖芯沉積物，發現有力 證據印證非洲乾旱氣候與人類發展的關係，例如氣候的轉 變逼使人類學習新的生活方式及發展新的技術。該項研 究成果已在國際權威跨學科期刊《美國國家科學院院刊》 刊登。 A research group led by Dr. Sung Yik Hei of the Department of Biology has described a new species of gecko in Hong Kong, and named it Hong Kong Tree Gecko. It is the eighth species of gecko and the only species of tree gecko found in Hong Kong. The species description was published in Zootaxa , a peer-reviewed scientific journal on the study of taxonomy. 生物系宋亦希博士率領的團隊發表了一種壁虎新品種，並 命名為「香港半葉趾虎」，是香港第八種壁虎，亦是唯一 一種半葉趾虎。是項發現已獲國際學術期刊《 Zootaxa 》 刊登。 Professor Chu Xiaowen and PhD student Shi Shaohuai from the Department of Computer Science have partnered with a team from Tencent Machine Learning to create a new technique for training Artificial Intelligence machines faster while maintaining accuracy. During the experiment the team succeeded in training two popular deep neural networks, AlexNet and ResNet-50, in just 4 and 6.6 minutes respectively. Previously the fastest times were 11 and 15 minutes. 計算機科學系褚曉文教授和博士生施少懷與騰訊機智機器 學習平台的團隊合作，創出一項嶄新方法，以更快速地教 導電腦學習辨別資料而不影響準確度。該方法分別只用了 4 分鐘和 6.6 分鐘便完成訓練 AlexNet 和 ResNet-50 兩個系統。過 往的國際紀錄分別是 11 分鐘和 15 分鐘。 Innovative Research 創新研究 18 University’s Recent Developments 大學最新發展