Zeta Hull Pursuits: Learning Nonconvex Data Hulls
Authors: Yuanjun Xiong, Wei Liu, Deli Zhao, Xiaoou Tang
NeurIPS 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experimental results show that data representation learned by Zeta Hulls can achieve state-of-the-art accuracy in text and image classification tasks. |
| Researcher Affiliation | Collaboration | Information Engineering Department, The Chinese University of Hong Kong, Hong Kong IBM T. J. Watson Research Center, Yorktown Heights, New York, USA Advanced Algorithm Research Group, HTC, Beijing, China |
| Pseudocode | Yes | Algorithm 1 Zeta Hull Pursuits |
| Open Source Code | No | The paper does not contain any explicit statements about releasing source code or links to a code repository for the described methodology. |
| Open Datasets | Yes | The MNIST dataset serves as a standard benchmark for machine learning algorithms. The Caltech101 dataset [17] is a widely used benchmark for object recognition systems. The Multi PIE human face dataset is a widely applied benchmark for face recognition [9]. |
| Dataset Splits | No | The paper specifies training and testing sets for various datasets (e.g., 'The training set has 60000 images and the testing set has 10000 images' for MNIST), but does not explicitly mention a validation set or specific validation split percentages. |
| Hardware Specification | No | The paper mentions that computations can be parallelized using a 'multi-core CPU or a modern GPU', but does not provide specific hardware details like model numbers or configurations. |
| Software Dependencies | No | The paper mentions using a 'standard ℓ1-regularized projection algorithm (LASSO)' and 'linear SVM' but does not specify any software names with version numbers. |
| Experiment Setup | Yes | In all experiments, the parameter z is fixed at 0.05 to guarantee the convergence of the Zeta function. For the A-ZHP algorithm, the parameter s is fixed at 10 and the number of anchor points l is set as 10% of the training set size. |