Uncertain Evidence in Probabilistic Models and Stochastic Simulators
Authors: Andreas Munk, Alexander Mead, Frank Wood
ICML 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | To showcase the impact of different interpretations of the same uncertain evidence, we carry out experiments in which one interpretation is defined as correct. We then compare inference results from each different interpretation illustrating the importance of careful consideration of uncertain evidence. In this section the importance of making the appropriate interpretation and treatment of uncertain evidence is illustrated. It contains three experiments constructed such that in experiment one and three, the appropriate treatment of the given uncertain evidence is to use Jeffrey s rule. |
| Researcher Affiliation | Collaboration | 1Department of Computer Science, University of British Columbia, Vancouver, B.C., Canada 2Inverted AI Ltd., Vancouver, B.C., Canada 3Mila, CIFAR AI Chair. |
| Pseudocode | No | The paper does not contain any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | No | The paper references 'PYPROB' (https://github.com/probprog/pyprob) as a tool used, but does not state that the code for the methodology described in this paper is open-source or available. |
| Open Datasets | Yes | Data is simulated based on Kepler-90g, with P = 210 days, e = 0.05, ω = 100 deg and ω + M = 198 deg using TTVFAST (Deck et al., 2014). The Kepler satellite (Borucki et al., 2010) measured the flux from over half a million stars over 5 years. |
| Dataset Splits | No | The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning. It uses simulated data for its experiments. |
| Hardware Specification | No | The paper mentions general computing resources ('West Grid', 'Compute Canada', and 'Advanced Research Computing at the University of British Columbia') but does not provide specific hardware details such as GPU/CPU models, processor types, or memory amounts used for running its experiments. |
| Software Dependencies | No | The paper mentions using PYPROB and TTVFAST but does not provide specific version numbers for these or any other software dependencies needed to replicate the experiments. |
| Experiment Setup | Yes | In all cases, the posteriors p(x|ζ) are Gaussian, with the different posteriors shown in Figure 2. Note how in Figure 2, in the left panel the three methods result in vastly different posteriors, whereas those in the right panel are indistinguishable. This emphasizes the importance of carefully choosing the approach in dealing with uncertain evidence. The prior over P is taken to be normal with 210 1 days. The prior over eccentricity is taken to be uniform between 0 and 0.15. The angular variables have uniform priors between 0 and 360 deg. Data is simulated based on Kepler-90g, with P = 210 days, e = 0.05, ω = 100 deg and ω + M = 198 deg using TTVFAST (Deck et al., 2014). |