Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Intrinsic Gaussian Process on Unknown Manifolds with Probabilistic Metrics
Authors: Mu Niu, Zhenwen Dai, Pokman Cheung, Yizhu Wang
JMLR 2023 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The applications of GPUM are illustrated in the simulation studies on the Swiss roll, high dimensional real datasets of Wi Fi signals and image data examples. Its performance is compared with the Graph Laplacian GP, Graph Mat ern GP and Euclidean GP. |
| Researcher Affiliation | Collaboration | Mu Niu EMAIL School of Mathematics and Statistics University of Glasgow, UK Zhenwen Dai EMAIL Spotify, London, UK Pokman Cheung EMAIL London, UK Yizhu Wang EMAIL School of Mathematics and Statistics University of Glasgow, UK |
| Pseudocode | Yes | Algorithm 1: Simulating BM sample paths on M for estimating Kt heat ... Algorithm 2: GL Algorithm. |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described in this paper. It mentions using a Github repository for a third-party implementation (GM-GP) for comparison, but not for their own GPUM method. |
| Open Datasets | Yes | The applications of GPUM are illustrated in the simulation studies on the Swiss roll, high dimensional real datasets of Wi Fi signals (Ferris et al., 2007) and image data examples (Nayar and Murase, 1996). |
| Dataset Splits | Yes | The point cloud is comprised of the set of labeled points n = 24 and the set of unlabeled points v = 450. ... Only n = 3 of the locations are labeled with one dimensional location coordinates of the mobile device and we aim at predicting the indoor locations of the remaining points based on the Wi Fi signals. ... n=13 images are randomly selected and used as the training set (labeled data). They are plotted as the red stars in the latent space in Fig.9(b). The remaining 53 images are used as the testing set (unlabeled data) which are plotted as the blue triangles in the latent space in Fig. 9(b). |
| Hardware Specification | No | The paper does not provide specific hardware details used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers used for its own implementation. It mentions using a third-party implementation, but not the specific tools or versions for their method. |
| Experiment Setup | Yes | The diffusion time is fixed at 50. ... Optimisation of the diffusion time t can be done by selecting the corresponding ÎŁt ff which maximizes the log marginal likelihood. Estimation of Ďh follows standard optimisation routines, such as quasi-Newton. NBM = 20000 BM paths are simulated. The BM transition density is evaluated using (25) at twenty nine target points {sj M R3|j {1, , 29}} in the observation space. |