A novel fusion strategy for E-MAP missing value imputation

作者：Pan, Xiao Yong; Tian, Ye; Yao, Li Xiu; Shen, Hong Bin

DOI：10.3182/20110828-6-IT-1002.01010

• Ensemble estimator
• Fusion approaches
• Fusion strategies
• Map data
• Missing value estimation
• Missing value imputation
• Missing values

Epistatic miniarray profiling (E-MAP) is powerful for measuring gene biological relevance. However, E-MAP suffers from large number of missing values, and in order to use the E-MAP information more efficiently, the missing values have to be estimated. In this paper, considering advantages and disadvantages of different independent algorithms, we proposed a novel fusion approach based on the high-level diversity to estimate missing values that consists of two global and four local base estimators. Experiment results show our fusion scheme is more effective and robust for the missing value imputations and outperforms all single base algorithms on E-MAP data.

• 出版日期2011
• 单位上海交通大学