Boosting is one of the most representational ensemble prediction methods. It can be divided into two se-ries: Boost-by-majority and Adaboost. This paper briefly introduces the research status of Boosting and one of it...Boosting is one of the most representational ensemble prediction methods. It can be divided into two se-ries: Boost-by-majority and Adaboost. This paper briefly introduces the research status of Boosting and one of its seri-als-AdaBoost,analyzes the typical algorithms of AdaBoost.展开更多
This paper discusses the ideas for designing modern information processing system,in which we emphasize the observation about the advantages and disadvantages while designing large-scale open system using MAS theory. ...This paper discusses the ideas for designing modern information processing system,in which we emphasize the observation about the advantages and disadvantages while designing large-scale open system using MAS theory. After brief review and analysis about the study situations or limitations of present information processing techniques and DAI/MAS,this paper discusses many problems in the combination between MAS and information processing. Finally,a brand new MAS organization structural model and new ideas for devising information processing systems are put forward with the consideration about the limitations of MAS' organization structure and the strong requirements of feasibility and efficiency.展开更多
使用分布式哈希表(distributed hash table,简称DHT)的应用系统必须在O(1)发现算法和O(logN)发现算法系列中选择适应的DHT协议.但是,不同网络波动程度的应用场景要求理想的DHT协议根据网络波动率能够自适应地调整.提出一种发现算法ROAD(...使用分布式哈希表(distributed hash table,简称DHT)的应用系统必须在O(1)发现算法和O(logN)发现算法系列中选择适应的DHT协议.但是,不同网络波动程度的应用场景要求理想的DHT协议根据网络波动率能够自适应地调整.提出一种发现算法ROAD(routing on active and demand),在延时和波动率之间自适应地调整以提供更好的性能.设计ROAD的关键挑战是构建路由表和降低延时的算法.通过构建加速路由表,加快发现服务的速度,降低消息转发的延时,并通过幂次序组播算法改善对超级点的依赖性.模拟实验显示,与现有DHT算法相比,ROAD维护了一种高效发现延时与波动率的折衷.选择不同质量类型的超级点,ROAD可以扩展成满足不同服务需要的发现机制.展开更多
文摘Boosting is one of the most representational ensemble prediction methods. It can be divided into two se-ries: Boost-by-majority and Adaboost. This paper briefly introduces the research status of Boosting and one of its seri-als-AdaBoost,analyzes the typical algorithms of AdaBoost.
文摘This paper discusses the ideas for designing modern information processing system,in which we emphasize the observation about the advantages and disadvantages while designing large-scale open system using MAS theory. After brief review and analysis about the study situations or limitations of present information processing techniques and DAI/MAS,this paper discusses many problems in the combination between MAS and information processing. Finally,a brand new MAS organization structural model and new ideas for devising information processing systems are put forward with the consideration about the limitations of MAS' organization structure and the strong requirements of feasibility and efficiency.
基金Supported by the National Natural Science Foundation of China under Grant Nos.60433040 60603070 (国家自然科学基金)the China Postdoctoral Science Foundation under Grant No.2005038064 (中国博士后基金)
文摘使用分布式哈希表(distributed hash table,简称DHT)的应用系统必须在O(1)发现算法和O(logN)发现算法系列中选择适应的DHT协议.但是,不同网络波动程度的应用场景要求理想的DHT协议根据网络波动率能够自适应地调整.提出一种发现算法ROAD(routing on active and demand),在延时和波动率之间自适应地调整以提供更好的性能.设计ROAD的关键挑战是构建路由表和降低延时的算法.通过构建加速路由表,加快发现服务的速度,降低消息转发的延时,并通过幂次序组播算法改善对超级点的依赖性.模拟实验显示,与现有DHT算法相比,ROAD维护了一种高效发现延时与波动率的折衷.选择不同质量类型的超级点,ROAD可以扩展成满足不同服务需要的发现机制.
