Kernel functions based on triplet comparisons
Authors: Matthäus Kleindessner, Ulrike von Luxburg
NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on both artificial and real data show that this is indeed the case and that the similarity scores defined by our kernel functions are meaningful. |
| Researcher Affiliation | Academia | Matthäus Kleindessner Department of Computer Science Rutgers University Piscataway, NJ 08854 mk1572@cs.rutgers.edu Ulrike von Luxburg Department of Computer Science University of Tübingen Max Planck Institute for Intelligent Systems, Tübingen luxburg@informatik.uni-tuebingen.de |
| Pseudocode | No | The paper describes the mathematical formulations of its kernel functions but does not provide pseudocode or a clearly labeled algorithm block. |
| Open Source Code | No | The paper mentions external code for comparison methods (e.g., t-STE) and data from previous work by the authors, but does not provide an explicit statement or link to the open-source code for the kernel functions proposed in this paper. |
| Open Datasets | Yes | The fifth data set consists of 400 vertices of an undirected graph from a stochastic block model and d equals the shortest path distance. We computed k1 and k2 based on 10% of all possible similarity triplets (chosen uniformly at random from all triplets). The results for the first two data sets are shown in Figure 3. The results for the remaining data sets are shown in Figure 6 in Section A.1 in the supplementary material. |
| Dataset Splits | No | The paper describes how points are selected (e.g., '1000 points uniformly at random') and how input triplets are generated, but does not provide explicit training, validation, or test dataset splits with percentages or counts for reproducibility. |
| Hardware Specification | Yes | All computations were performed in Matlab R2016a on a Mac Book Pro with 2.9 GHz Intel Core i7 and 8 GB 1600 MHz DDR3. |
| Software Dependencies | Yes | All computations were performed in Matlab R2016a |
| Experiment Setup | Yes | Collections S of similarity triplets were generated as follows: We chose a certain number of landmark objects uniformly at random from all objects of the data set under consideration. Choosing d as the Euclidean metric, we created answers to all possible distance comparisons with the landmark objects as explained in Section 2.3. Answers were incorrect with some probability ep 1 independently of each other. From the set of all answers we chose triplets in S uniformly at random without replacement. |