Axel Gandy, Xing Liu, Andrew Duncan
We lifted 16 functions out of this paper's own repositories and ran 0 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| XingLLiu/pksd | canonical | 0 of 16 |
| Function | Status | Where it lives |
|---|---|---|
| compute_ksd | Not yet run | XingLLiu/pksd/pksd/kgof/kernel.py code served (permissive licence) · get_code("48932edc3b597c09") |
| create_mixture_gaussian_kdim_logprobb | Not yet run | XingLLiu/pksd/pksd/models_np.py code served (permissive licence) · get_code("703b535ace0bba4b") |
| ksdagg_wild_custom | Not yet run | XingLLiu/pksd/pksd/kgof/ksdagg.py code served (permissive licence) · get_code("01ab641d9bba2730") |
| l2norm | Not yet run | XingLLiu/pksd/pksd/kernel.py code served (permissive licence) · get_code("5ac1129cecb37805") |
| logsumexp | Not yet run | XingLLiu/pksd/pksd/models_np.py code served (permissive licence) · get_code("1e9f4edd607a92bd") |
| merge_modes | Not yet run | XingLLiu/pksd/pksd/find_modes.py code served (permissive licence) · get_code("0e4dfe17d8b7d807") |
| multivariate_t_logprob | Not yet run | XingLLiu/pksd/pksd/models_np.py code served (permissive licence) · get_code("6a15096c15f4ca4e") |
| norm2_sq | Not yet run | XingLLiu/pksd/pksd/sensors.py code served (permissive licence) · get_code("b3810f48851c8037") |
| norm2_sq_np | Not yet run | XingLLiu/pksd/pksd/sensors.py code served (permissive licence) · get_code("c5325727ede27f6e") |
| pairwise_mahalanobis | Not yet run | XingLLiu/pksd/pksd/find_modes.py code served (permissive licence) · get_code("16fc88a882225eb1") |
| plot_sensors | Not yet run | XingLLiu/pksd/pksd/sensors.py code served (permissive licence) · get_code("fade3727070787f0") |
| prepare_proposal_input | Not yet run | XingLLiu/pksd/pksd/langevin.py code served (permissive licence) · get_code("44916ebc94da1c8f") |
| prepare_proposal_input_all | Not yet run | XingLLiu/pksd/pksd/langevin.py code served (permissive licence) · get_code("f0eef89f6029121e") |
| ratio_ksd_stdev | Not yet run | XingLLiu/pksd/pksd/kgof/kernel.py code served (permissive licence) · get_code("afdccdc29f702ebf") |
| stein_kernel_matrices | Not yet run | XingLLiu/pksd/pksd/kgof/kernel.py code served (permissive licence) · get_code("ec8a35d4a6925338") |
| t_prob_fast | Not yet run | XingLLiu/pksd/pksd/models.py code served (permissive licence) · get_code("94592dc674f4f6a4") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Kernelized Stein discrepancy (KSD) is a scorebased discrepancy widely used in goodness-of-fit tests. It can be applied even when the target distribution has an unknown normalising factor, such as in Bayesian analysis. We show theoretically and empirically that the KSD test can suffer from low power when the target and the alternative distributions have the same well-separated modes but differ in mixing proportions. We propose to perturb the observed sample via Markov transition kernels, with respect to which the target distribution is invariant. This allows us to then employ the KSD test on the perturbed sample. We provide numerical evidence that with suitably chosen transition kernels the proposed approach can lead to substantially higher power than the KSD test.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2304.14762")
get_code_for_paper("2304.14762")
have("2304.14762")
Connect an agent — have() is free.