Robust Ordinal Embedding from Contaminated Relative Comparisons
Authors: Ke Ma, Qianqian Xu, Xiaochun Cao7908-7915
AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our studies are supported by experiments with both simulated examples and real-world data. The results demonstrate that our method outperforms the state-of-the-art alternatives. |
| Researcher Affiliation | Academia | 1State Key Laboratory of Information Security, Institute of Information Engineering, Chinese Academy of Sciences 2School of Cyber Security, University of Chinese Academy of Sciences 3Key Laboratory of Intelligent Information Processing, Institute of Computing Technology, Chinese Academy of Sciences |
| Pseudocode | Yes | Algorithm 1: rank-reduction(β+, p) and Algorithm 2: FISTA with rank reduction for (14) |
| Open Source Code | Yes | The reproducible code can be found here2. 2https://github.com/alphaprime/ROE |
| Open Datasets | Yes | The music artist data is collected by (Ellis et al. 2002) via a web-based survey... |
| Dataset Splits | Yes | We randomly sample 10000 correct triplets as the basic training data and a validation set is build with the same number of triplets. The remains are served as the test set. The size of training samples is 5, 000 and the validation set contains 2, 000 triplets. The rest of triplets are treated as test set. |
| Hardware Specification | No | The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments. It does not specify the machine where the computations were performed. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers, such as programming language versions or library versions (e.g., Python 3.8, PyTorch 1.9). |
| Experiment Setup | Yes | The desired dimension of embedding is d = 9 as these music artists can be classified by genre into 9 categories. For each triplet t, there will be s copies, t1, . . . , ts, 15 s 50. errors are then synthesized according to the different ratios: we assume that at most q% of all relative comparisons are not consistent with the ground-truth metric information and we change the position of j and k in each randomly chosen triplet, where q ranges from 10% to 25%. |