Fast and Flexible ADMM Algorithms for Trend Filtering

@article{Ramdas2014FastAF,
  title={Fast and Flexible ADMM Algorithms for Trend Filtering},
  author={Aaditya Ramdas and Ryan J. Tibshirani},
  journal={CoRR},
  year={2014},
  volume={abs/1406.2082}
}
This paper presents a fast and robust algorithm for trend filtering, a recently developed nonparametric regression tool. It has been shown that, for estimating functions whose derivatives are of bounded variation, trend filtering achieves the minimax optimal error rate, while other popular methods like smoothing splines and kernels do not. Standing in the way of a more widespread practical adoption, however, is a lack of scalable and numerically stable algorithms for fitting trend filtering… CONTINUE READING
Highly Cited
This paper has 141 citations. REVIEW CITATIONS
Recent Discussions
This paper has been referenced on Twitter 2 times over the past 90 days. VIEW TWEETS
25 Citations
19 References
Similar Papers

Citations

Publications citing this paper.

141 Citations

02040608020142015201620172018
Citations per Year
Semantic Scholar estimates that this publication has 141 citations based on the available data.

See our FAQ for additional information.

References

Publications referenced by this paper.
Showing 1-10 of 19 references

A Practical Guide to Splines

  • C. de Boor
  • 1978
Highly Influential
3 Excerpts

A dynamic programming algorithm for the fused lasso and L 0segmentation ’

  • N. Johnson
  • Journal of Computational and Graphical Statistics
  • 2013
Highly Influential
3 Excerpts

Coordinate descent, on the other hand, is quite far off (although we note that it does deliver a visually perfect piecewise linear fit after nearly 100,000 iterations)

  • Kim
  • 2009
Highly Influential
2 Excerpts

Rn×n is kth order falling factorial basis matrix, defined over

  • H H
  • 2014

argue that the appropriate extension of f̂ to the continuous domain is given by f̂(x

  • β̂n. Tibshirani, Wang
  • 2014

Similar Papers

Loading similar papers…