simulation {muscor}R Documentation

Simulated high dimensional regression problem with sparse target

Description

This data is generated according to the simulation experiments described in [Tong Zhang (2008)].

Format

The list simulation contains the following components:

x
a 100 x 500 matrix
y
a 100 dimensional vector
true.coeff
a 500 dimension vector of the true coefficients
true.set
a 5 dimension vector indicating the true feature set

Details

The data contains n=100 training examples, with p=500 dimensions. The true regression coefficient vector is sparse, with only five nonzero coefficients.

References

Tong Zhang (2008) "Adaptive Forward-Backward Greedy Algorithm for Learning Sparse Representations", Rutgers Technical Report (long version).

Tong Zhang (2008) "Adaptive Forward-Backward Greedy Algorithm for Sparse Learning with Linear Models", NIPS'08 (short version).


[Package muscor version 0.2 Index]