These instances, whenmapped to an N-dimensional space, represent a core set that can be
used to construct an approximation to theminimumenclosing ball. Solving the SVMlearning
problem on these core sets can produce a good approximation solution in very fast speed.
For example, the core-vector machine [81] thus produced can learn an SVM for millions of
data in seconds.
基于T i n y O S 的無線傳感器網(wǎng)絡(luò)體系結(jié)構(gòu)
無線傳感器網(wǎng)絡(luò)WSN(Wireless Sensor Network) 由部署在監(jiān)測區(qū)域內(nèi)大量廉價(jià)的傳感器節(jié)點(diǎn)組成,通過
無線通信方式形成multi2hop 自組織的網(wǎng)絡(luò)系統(tǒng),其目的是協(xié)作地感知、采集和處理網(wǎng)絡(luò)覆蓋區(qū)域中感知
對象的信息,并發(fā)送給觀察者。本文從分析無線傳感器節(jié)點(diǎn)Mica 和其上運(yùn)行的操作系統(tǒng)TinyOS 出發(fā),
著重描述無線傳感器網(wǎng)絡(luò)節(jié)點(diǎn)應(yīng)用程序體系結(jié)構(gòu)和消息通信機(jī)制。
