High-Dimensional Sparse Additive Hazards Regression
88 citations recorded by the source corpus.
High-dimensional sparse modeling with censored survival data is of great practical importance, as exemplified by modern applications in high-throughput genomic data analysis and credit risk analysis. In this article, we propose a class of regularization methods for simultaneous variable selection and estimation in the additive hazards model, by combining the nonconcave penalized likelihood approach and the pseudoscore method. In a high-dimensional setting where the dimensionality can grow fast, polynomially or nonpolynomially, with the sample size, we establish the weak oracle property and oracle property under mild, interpretable conditions, thus providing strong performance guarantees for the proposed methodology. Moreover, we show that the regularity conditions required by the L 1 method are substantially relaxed by a certain class of sparsity-inducing concave penalties. As a result, concave penalties such as the smoothly clipped absolute deviation, minimax concave penalty, and smooth integration of counting and absolute deviation can significantly improve on the L 1 method and yield sparser models with better prediction performance. We present a coordinate descent algorithm for efficient implementation and rigorously investigate its convergence properties. The practical use and effectiveness of the proposed methods are demonstrated by simulation studies and a real data example.
Subfields: Artificial Intelligence, Genetics, Statistics and Probability
View Paper: https://www.semanticscholar.org/paper/cd6bf4b3c2eeab4e030d40954fe9c300dc22d5f5
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
In Genetics
In Statistics and Probability
Citations in the corpus
- Asymptotic Equivalence of Regularization Methods in Thresholded Parameter Space (2013, methodology): “om the previous step along that coordinate reduces the problem to a univariate penalized least squares, which admits a closed-form solution for many commonly used penalty functions. See, for example, Lin and Lv (2013) for an analysis of convergence properties of this algorithm. In this paper, we app”
- A General Theory of Hypothesis Tests and Confidence Regions for Sparse High Dimensional Models (2014, unknown): “To estimate β , [18, 23] proposed the estimator β̂ in (1.”
- High dimensional thresholded regression and shrinkage effect (2014, methodology): “See also Lin and Lv (2013) for the closed-form solution of the univariate SICA estimator.”