We lifted 9 functions out of this paper's own repositories and ran 9 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 |
|---|---|---|
| willieneis/gp-martingales | pwc_unofficial | 9 of 9 |
| Function | Status | Where it lives |
|---|---|---|
| branin | Ran | willieneis/gp-martingales/src/branin.py code served (permissive licence) · get_code("7437e0f18c2fdb52") |
| branin_on_matrix | Ran | willieneis/gp-martingales/src/branin.py code served (permissive licence) · get_code("dd191a4fa4ed3f21") |
| branin_single | Ran | willieneis/gp-martingales/src/branin.py code served (permissive licence) · get_code("c2531c541561e89f") |
| dict_to_namespace | Ran | willieneis/gp-martingales/src/simple_gp.py code served (permissive licence) · get_code("8a3545000b23b4c3") |
| get_data_from_sample_path | Ran | willieneis/gp-martingales/src/gp_util.py code served (permissive licence) · get_code("ced0f63b21189618") |
| get_noisy_sample_path | Ran | willieneis/gp-martingales/src/gp_util.py code served (permissive licence) · get_code("89ca0dbdec25a646") |
| get_sample_path | Ran | willieneis/gp-martingales/src/gp_util.py code served (permissive licence) · get_code("ca0287c7d6d9a2aa") |
| normal_quotient | Ran | willieneis/gp-martingales/src/quotient.py code served (permissive licence) · get_code("6b4a712b6acac1eb") |
| normal_quotient_log_z | Ran | willieneis/gp-martingales/src/quotient.py code served (permissive licence) · get_code("b0b4e1d7e5cb31cd") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
We address uncertainty quantification for Gaussian processes (GPs) under misspecified priors, with an eye towards Bayesian Optimization (BO). GPs are widely used in BO because they easily enable exploration based on posterior uncertainty bands. However, this convenience comes at the cost of robustness: a typical function encountered in practice is unlikely to have been drawn from the data scientist's prior, in which case uncertainty estimates can be misleading, and the resulting exploration can be suboptimal. We present a frequentist approach to GP/BO uncertainty quantification. We utilize the GP framework as a working model, but do not assume correctness of the prior. We instead construct a confidence sequence (CS) for the unknown function using martingale techniques. There is a necessary cost to achieving robustness: if the prior was correct, posterior GP bands are narrower than our CS. Nevertheless, when the prior is wrong, our CS is statistically valid and empirically outperforms standard GP methods, in terms of both coverage and utility for BO. Additionally, we demonstrate that powered likelihoods provide robustness against model misspecification.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2006.07368")
get_code_for_paper("2006.07368")
have("2006.07368")
Connect an agent — have() is free.