Robust Principal Component Analysis with Side Information

Authors: Kai-Yang Chiang, Cho-Jui Hsieh, Inderjit Dhillon

ICML 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In addition, we conduct synthetic experiments as well as a real application on noisy image classification to show that our method also improves the performance in practice by exploiting side information.
Researcher Affiliation Academia Kai-Yang Chiang? KYCHIANG@CS.UTEXAS.EDU Cho-Jui Hsieh CHOHSIEH@UCDAVIS.EDU Inderjit S. Dhillon? INDERJIT@CS.UTEXAS.EDU ?Department of Computer Science, The University of Texas at Austin, Austin, TX 78712, USA Department of Statistics and Computer Science, University of California at Davis, Davis, CA 95616, USA
Pseudocode Yes Algorithm 1 ALM method for PCPF
Open Source Code No The paper mentions using a third-party ALM solver for PCP but does not provide any link or statement about the availability of their own code for PCPF.
Open Datasets Yes We consider the digit recognition dataset MNIST, which includes 50,000 training images and 10,000 testing images, and each image is a handwriting digit described as a 784 dimensional vector.
Dataset Splits No The paper mentions training and testing sets for MNIST but does not specify a validation set or explicit split percentages for all three.
Hardware Specification No The paper does not provide specific details about the hardware used for experiments.
Software Dependencies No The paper mentions using LIBLINEAR and LIBSVM with citations but does not provide specific version numbers for these or any other software dependencies.
Experiment Setup Yes The parameters λ in both PCP and PCPF are set as 1/pn by default as Theorem 1 suggested. The convergence criterion is set to be k R S XHY T k F /k Rk F < 10 7 as suggested in Cand es et al. (2011).