JUMP-Means: Small-Variance Asymptotics for Markov Jump Processes

Authors: Jonathan Huggins, Karthik Narasimhan, Ardavan Saeedi, Vikash Mansinghka

ICML 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our experiments demonstrate that JUMP-means is competitive with or outperforms widely used MJP inference approaches in terms of both speed and reconstruction accuracy.
Researcher Affiliation Academia Jonathan H. Huggins* JHUGGINS@MIT.EDU Karthik Narasimhan* KARTHIKN@MIT.EDU Ardavan Saeedi* ARDAVANS@MIT.EDU Vikash K. Mansinghka VKM@MIT.EDU Computer Science and Artificial Intelligence Laboratory, MIT
Pseudocode No The paper describes algorithms in numbered step-by-step prose within sections titled 'Algorithm' (e.g., Section 3.3, 4.1), but does not present them in a structured pseudocode block format.
Open Source Code No The paper does not provide concrete access to source code for the methodology described.
Open Datasets Yes For our experiments, we use a real-world dataset collected from a phase III clinical trial of a drug for MS. (Mandel, 2010). We use data from the MIMIC database (Goldberger et al., 2000; Moody & Mark, 1996)
Dataset Splits No The paper mentions holding out data for testing but does not explicitly specify a separate validation dataset split.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU/GPU models, memory) used for running its experiments.
Software Dependencies No The paper mentions implementation languages like 'Java' and 'Python' but does not provide specific version numbers for them or any other software dependencies.
Experiment Setup Yes We set the hyperparameters ξ, ξλ, and µλ equal to 1, 1, and .5, respectively. (Section 5.1); The hyperparameters γ and ξ1 are set to 5, while ζ, ξ, and ξ2 are set to 0.005. (Section 5.2)