Robust Inverse Covariance Estimation under Noisy Measurements

Jun-Kun Wang, Shou-de Lin
Proceedings of the 31st International Conference on Machine Learning, PMLR 32(2):928-936, 2014.

Abstract

This paper proposes a robust method to estimate the inverse covariance under noisy measurements. The method is based on the estimation of each column in the inverse covariance matrix independently via robust regression, which enables parallelization. Different from previous linear programming based methods that cannot guarantee a positive semi-definite covariance matrix, our method adjusts the learned matrix to satisfy this condition, which further facilitates the tasks of forecasting future values. Experiments on time series prediction and classification under noisy condition demonstrate the effectiveness of the approach.

Cite this Paper


BibTeX
@InProceedings{pmlr-v32-wangf14, title = {Robust Inverse Covariance Estimation under Noisy Measurements}, author = {Wang, Jun-Kun and Lin, Shou-de}, booktitle = {Proceedings of the 31st International Conference on Machine Learning}, pages = {928--936}, year = {2014}, editor = {Xing, Eric P. and Jebara, Tony}, volume = {32}, number = {2}, series = {Proceedings of Machine Learning Research}, address = {Bejing, China}, month = {22--24 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v32/wangf14.pdf}, url = {https://proceedings.mlr.press/v32/wangf14.html}, abstract = {This paper proposes a robust method to estimate the inverse covariance under noisy measurements. The method is based on the estimation of each column in the inverse covariance matrix independently via robust regression, which enables parallelization. Different from previous linear programming based methods that cannot guarantee a positive semi-definite covariance matrix, our method adjusts the learned matrix to satisfy this condition, which further facilitates the tasks of forecasting future values. Experiments on time series prediction and classification under noisy condition demonstrate the effectiveness of the approach.} }
Endnote
%0 Conference Paper %T Robust Inverse Covariance Estimation under Noisy Measurements %A Jun-Kun Wang %A Shou-de Lin %B Proceedings of the 31st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2014 %E Eric P. Xing %E Tony Jebara %F pmlr-v32-wangf14 %I PMLR %P 928--936 %U https://proceedings.mlr.press/v32/wangf14.html %V 32 %N 2 %X This paper proposes a robust method to estimate the inverse covariance under noisy measurements. The method is based on the estimation of each column in the inverse covariance matrix independently via robust regression, which enables parallelization. Different from previous linear programming based methods that cannot guarantee a positive semi-definite covariance matrix, our method adjusts the learned matrix to satisfy this condition, which further facilitates the tasks of forecasting future values. Experiments on time series prediction and classification under noisy condition demonstrate the effectiveness of the approach.
RIS
TY - CPAPER TI - Robust Inverse Covariance Estimation under Noisy Measurements AU - Jun-Kun Wang AU - Shou-de Lin BT - Proceedings of the 31st International Conference on Machine Learning DA - 2014/06/18 ED - Eric P. Xing ED - Tony Jebara ID - pmlr-v32-wangf14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 32 IS - 2 SP - 928 EP - 936 L1 - http://proceedings.mlr.press/v32/wangf14.pdf UR - https://proceedings.mlr.press/v32/wangf14.html AB - This paper proposes a robust method to estimate the inverse covariance under noisy measurements. The method is based on the estimation of each column in the inverse covariance matrix independently via robust regression, which enables parallelization. Different from previous linear programming based methods that cannot guarantee a positive semi-definite covariance matrix, our method adjusts the learned matrix to satisfy this condition, which further facilitates the tasks of forecasting future values. Experiments on time series prediction and classification under noisy condition demonstrate the effectiveness of the approach. ER -
APA
Wang, J. & Lin, S.. (2014). Robust Inverse Covariance Estimation under Noisy Measurements. Proceedings of the 31st International Conference on Machine Learning, in Proceedings of Machine Learning Research 32(2):928-936 Available from https://proceedings.mlr.press/v32/wangf14.html.

Related Material