Uncorrelated Group LASSO

Authors: Deguang Kong, Ji Liu, Bo Liu, Xuan Bao

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experiment results on real world datasets demonstrate the effectiveness of the proposed new regularization and algorithm. and To validate the effectiveness of our method, we conduct experiments on multi-label datasets for image annotation.
Researcher Affiliation Collaboration Deguang Kong1, Ji Liu2, Bo Liu3 and Xuan Bao4 1Samsung Research America, 2University of Rochester, 3Philips Research North America, 4Google Inc.
Pseudocode No The paper describes the optimization algorithm through text and mathematical equations (Eq. 11, 12) and iterative steps, but does not present it as a clearly labeled pseudocode or algorithm block/figure.
Open Source Code No The paper does not provide any specific repository links, explicit code release statements, or mention code in supplementary materials for the described methodology.
Open Datasets Yes Take the House dataset1 (n=506, p=14) as an example. ... 1http://archive.ics.uci.edu/ml/datasets/housing and Barcelona5 dataset contains... 5http://mlg.ucd.ie/content/view/61 and MSRC6 dataset contains... 6http://research.microsoft.com/en-us/projects/\ objectclassrecognition/ and TREVID20057 dataset contains... 7http://www-nlpir.nist.gov/projects/tv2005/
Dataset Splits Yes In all the experiments, we use 5-fold cross validation, where 4-fold data are used for training and the remaining ones are used for testing purpose.
Hardware Specification No The paper does not provide any specific hardware details such as GPU/CPU models, processor types, or memory amounts used for running experiments.
Software Dependencies No The paper mentions various software concepts and methods, but does not provide specific version numbers for any ancillary software or libraries used in its experiments.
Experiment Setup Yes In our approach of Eq.(2), we use logistic loss function, group Gg is generated according to the feature correlation defined in Eq.(9), where θ = 0.3. and We adjust the parameter α such that the number of nonzero rows in W (i.e., optimal solution) is r.