Indexed by:
Abstract:
为提高文本分类的准确率,提出了费希尔信息度量随机近邻嵌入算法(Fisher information metric based on stochastic neighbor embedding,FIMSNE).首先,把文本的词频向量看作统计流形上的概率密度样本点,利用费希尔信息度量计算样本点之间的距离;然后,从信息几何的观点出发,对t分布随机近邻嵌入(t-stochastic neighbor embedding,t-SNE)进行改进,实现了新算法.真实文本数据集上的二维嵌入和分类实验的结果表明:FIMSNE的性能在总体上优于t-SNE、费希尔信息非参数嵌入(Fisher information...
Keyword:
Reprint Author's Address:
Email:
Source :
北京工业大学学报
Year: 2016
Issue: 06
Volume: 42
Page: 862-869
Cited Count:
WoS CC Cited Count: 0
SCOPUS Cited Count:
ESI Highly Cited Papers on the List: 0 Unfold All
WanFang Cited Count:
Chinese Cited Count:
30 Days PV: 1