Efficient Approximation of Cross-Validation for Kernel Methods using Bouligand Influence Function

Authors: Yong Liu, Shali Jiang, Shizhong Liao

ICML 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results demonstrate that our approximate cross-validation criterion is sound and efficient.
Researcher Affiliation Academia Yong Liu YONGLIU@TJU.EDU.CN Shali Jiang SLJIANG@TJU.EDU.CN Shizhong Liao SZLIAO@TJU.EDU.CN School of Computer Science and Technology, Tianjin University, Tianjin 300072, P. R. China
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any concrete access to source code for the methodology described.
Open Datasets Yes The evaluation is made on 20 publicly available data sets from LIBSVM Data: 10 data sets for classification and 10 data sets for regression seen in Table 1.
Dataset Splits Yes For each data set, we have run all the methods 10 times with training and testing data sets be split randomly (50% of all the examples for training and the other 50% for testing). ... For each training set, we choose the τ and λ by cross validation on the training set.
Hardware Specification Yes Experiments are performed on a Dell Vestro PC with 3.4-GHz 8-core CPU and 8-GB memory.
Software Dependencies No The paper does not provide specific software names with version numbers.
Experiment Setup Yes We use K(x, x ) = exp( x x 2 2/2τ) as our candidate kernels, τ {2i, i = 6, 5, . . . , 7, 8} 2. The regularization parameter λ {2i, i = 7, 6, . . . , 2}.