A0866
Title: Safe feature identification rule for fused Lasso by an extra dual variable
Authors: Pan Shang - Academy of Mathematics and Systems Science, Chinese Academy of Sciences (China) [presenting]
Huangyue Chen - Chinese Academy of Sciences (China)
Lingchen Kong - Beijing Jiaotong University (China)
Abstract: Fused Lasso was proposed to characterize the sparsity of the coefficients and the sparsity of their successive differences for the linear regression. Due to its wide applications, there are many existing algorithms to solve fused Lasso. However, the computation of this model is time-consuming in high-dimensional data sets. To accelerate the calculation of fused Lasso in high-dimension data sets, the safe feature identification rule is built up by introducing an extra dual variable. With a low computational cost, this rule can eliminate inactive features with zero coefficients and identify adjacent features with the same coefficients in the solution. To the best of our knowledge, existing screening rules cannot be applied to speed up the computation of fused Lasso, and this is the first time to deal with this problem. To emphasize, the rule is a unique result that is capable of identifying adjacent features with the same coefficients; the result is named the safe feature identification rule. Numerical experiments on simulation and real data illustrate the efficiency of the rule, which means this rule can reduce the computational time of fused Lasso. In addition, the rule can be embedded into any efficient algorithm to speed up the computational process of fused Lasso.
