演講公告
新聞標題： ( 2017-06-05 )
演講主題：Smart Continuous Sensing and Monitoring
主講人：鮑興國教授(台科大資訊工程系)
演講日期：2017年6月6日(星期二) 14:00 – 15:00
演講地點：(光復校區) 科學一館223室
茶會時間：當天下午13:30 (科學一館205室)
摘要內容：
Abstract:
We utilize sensors to help us monitor events in the environment around us. In recent years, sensors
become “small machines”, sometimes built with various devices or equipment (called Things) so that other than considering a sensor network, we have a more powerful machine network or Internet of things to manage. In the cases of intelligent transportation systems, smart homes, or weather prediction, homogeneous or heterogeneous sensory machines are deployed on things or deployed on their own to constantly sense the environment and we can use the collected readings to detect events or anomalies for further investigation. From the data analytics viewpoint, common issues in the sensory machine network or Internet of things include how to deal with tons of data (BigData) that are generated by sensors, how to save power consumption or transmission bandwidth in the network, and how to deal with data missing or forgery, to name a few. We aim to propose a framework to deal with sensory machine or IoT data to make a smart sensory machines network possible. We can use a limited amount of sensing and at the same time the sensing is effective enough to “understand” the environment without too much knowledge loss. Another important feature in our goal is to take the smart sensing action continuously. We consider the smart sensing in an online fashion so that we can analyze the data continuously to give output about environmental status at any time. To achieve that, the methodology must efficient enough so that we can continuously given reasonable result given massive data for each moment. We discuss the sensory machine network in various scales, with various sampling rate, and the sensory machines may be homogeneous or heterogeneous such that the data are with different value ranges or owning different physical meanings. Some case studies shall be focused to illustrate the effectiveness and efficiency of the proposed methodology.相關檔案：Talk_20170605.pdf
go back