为了改善传统Relief算法适应性和鲁棒性差的缺陷,融合间距最大化、信息熵和分类局部一致性,构造了新的间距最大化目标函数,并进一步对目标函数进行优化,得到一些新的理论结果。在此基础上提出了新的基于两类数据的Relief特征加权算法LIE...为了改善传统Relief算法适应性和鲁棒性差的缺陷,融合间距最大化、信息熵和分类局部一致性,构造了新的间距最大化目标函数,并进一步对目标函数进行优化,得到一些新的理论结果。在此基础上提出了新的基于两类数据的Relief特征加权算法LIE-Relief-T(Local consistency information entropy Relief algorithm based twoclass data),并将其扩展到多类数据的特征加权算法LIE-Relief-MLocal consistency information entropy Relief algorithm based multi-class data)。利用UCI和基因表达数据集进行实验验证,结果表明该新的Relief特征加权算法分类错误率较低,对噪声和野点表现出了更好的适应性和鲁棒性。展开更多
Locally advanced pancreatic cancer is associated with a very poor prognosis. This study was performed to evaluate whether patients with locally advanced pancreatic cancer benefit from 125 ^I seed implantation. This re...Locally advanced pancreatic cancer is associated with a very poor prognosis. This study was performed to evaluate whether patients with locally advanced pancreatic cancer benefit from 125 ^I seed implantation. This retrospective study included 224 patients with locally advanced pancreatic cancer, with 137 patients（61.2%） in the implantation（IP） group and 87（38.9%） in the non-implantation（NIP） group. The survival status, complications and objective curative effects were compared between the groups. The average operative time in the IP group was significantly longer than that in the NIP group（243±51 vs. 214±77 min）. The tumor response rates were 9.5% and 0 at the 2nd month after surgery in the IP and NIP groups, respectively（P〈0.05）. The IP group exhibited a trend toward pain relief at the 6th month after surgery. The global health status scores of the IP group were higher than those of the NIP group at the 3rd and 6th month after surgery. The median survival time in the IP group was significantly longer than that in the NIP group. In conclusion, patients with locally advanced pancreatic cancer can benefit from 125 I seed implantation in terms of local tumor control, survival time, pain relief and quality of life.展开更多
文摘为了改善传统Relief算法适应性和鲁棒性差的缺陷,融合间距最大化、信息熵和分类局部一致性,构造了新的间距最大化目标函数,并进一步对目标函数进行优化,得到一些新的理论结果。在此基础上提出了新的基于两类数据的Relief特征加权算法LIE-Relief-T(Local consistency information entropy Relief algorithm based twoclass data),并将其扩展到多类数据的特征加权算法LIE-Relief-MLocal consistency information entropy Relief algorithm based multi-class data)。利用UCI和基因表达数据集进行实验验证,结果表明该新的Relief特征加权算法分类错误率较低,对噪声和野点表现出了更好的适应性和鲁棒性。
文摘Locally advanced pancreatic cancer is associated with a very poor prognosis. This study was performed to evaluate whether patients with locally advanced pancreatic cancer benefit from 125 ^I seed implantation. This retrospective study included 224 patients with locally advanced pancreatic cancer, with 137 patients（61.2%） in the implantation（IP） group and 87（38.9%） in the non-implantation（NIP） group. The survival status, complications and objective curative effects were compared between the groups. The average operative time in the IP group was significantly longer than that in the NIP group（243±51 vs. 214±77 min）. The tumor response rates were 9.5% and 0 at the 2nd month after surgery in the IP and NIP groups, respectively（P〈0.05）. The IP group exhibited a trend toward pain relief at the 6th month after surgery. The global health status scores of the IP group were higher than those of the NIP group at the 3rd and 6th month after surgery. The median survival time in the IP group was significantly longer than that in the NIP group. In conclusion, patients with locally advanced pancreatic cancer can benefit from 125 I seed implantation in terms of local tumor control, survival time, pain relief and quality of life.
