GSelection (version 0.1.0)

RED: Redundancy Rate

Description

Calculate the redundancy rate of the selected features(markers). Value will be high if many redundant features are selected.

Usage

RED(x,spam_selected_feature_index,hsic_selected_feature_index,
integrated_selected_feature_index)

Arguments

x

a matrix of markers or explanatory variables, each column contains one marker and each row represents an individual.

spam_selected_feature_index

index of selected markers from x using Sparse Additive Model.

hsic_selected_feature_index

index of selected markers from x using HSIC LASSO.

integrated_selected_feature_index

index of selected markers from x using integrated model framework

Value

Returns a LIST containing

RED_spam

returns redundancy rate of features selected by using Sparse Additive Model.

RED_hsic

returns redundancy rate of features selected by using HSIC LASSO.

RED_I

returns redundancy rate of features selected by using integrated model framework.

Details

The RED score (Zhao et al., 2010) is determined by average of the correlation between each pair of selected markers. A large RED score signifies that selected features are more strongly correlated to each other which means many redundant features are selected. Thus, a small redundancy rate is preferable for feature selection.

References

Guha Majumdar, S., Rai, A. and Mishra, D. C. (2019). Integrated framework for selection of additive and non-additive genetic markers for genomic selection. Journal of Computational Biology. doi:10.1089/cmb.2019.0223 Ravikumar, P., Lafferty, J., Liu, H. and Wasserman, L. (2009). Sparse additive models. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 71(5), 1009-1030. doi:10.1111/j.1467-9868.2009.00718.x Yamada, M., Jitkrittum, W., Sigal, L., Xing, E. P. and Sugiyama, M. (2014). High-Dimensional Feature Selection by Feature-Wise Kernelized Lasso. Neural Computation, 26(1):185-207. doi:10.1162/NECO_a_00537 Zhao, Z., Wang, L. and Li, H. (2010). Efficient spectral feature selection with minimum redundancy. In AAAI Conference on Artificial Intelligence (AAAI), pp 673-678.

Examples

# NOT RUN {
library(GSelection)
data(GS)
x_trn <- GS[1:40,1:110]
y_trn <- GS[1:40,111]
x_tst <- GS[41:60,1:110]
y_tst <- GS[41:60,111]
fit <- feature.selection(x_trn,y_trn,d=10)
red <- RED(x_trn,fit$spam_selected_feature_index,fit$hsic_selected_feature_index,
fit$integrated_selected_feature_index)
# }