SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2006.11615 · ICML · 2020

Scalable Identification of Partially Observed Systems with Certainty-Equivalent EM

Mykel Kochenderfer, Kunal Menda, Jean De Becdelièvre, Jayesh Gupta, Ilan Kroo, Zachary Manchester

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 10 functions out of this paper's own repositories and ran 8 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.

RepositoryRoleRan
sisl/CEEM canonical 8 of 10
FunctionStatusWhere it lives
HarmonicDecayScheduler Ran sisl/CEEM/ceem/particleem.py
code served (permissive licence) · get_code("b9784ff84cd8a51d")
compute_rms Ran sisl/CEEM/ceem/exp_utils.py
code served (permissive licence) · get_code("68918bf29793f782")
ensure_default_torch_kwargs Ran sisl/CEEM/ceem/learner.py
code served (permissive licence) · get_code("50f8cbf6de31a1af")
gen_ypred_benchmark Ran sisl/CEEM/ceem/exp_utils.py
code served (permissive licence) · get_code("15c3ab8d13c9c4ae")
load_helidata Ran sisl/CEEM/ceem/data_utils.py
code served (permissive licence) · get_code("b250895ef600807a")
load_statistics Ran sisl/CEEM/ceem/data_utils.py
code served (permissive licence) · get_code("05ede9c21387428f")
make_output_format Ran sisl/CEEM/ceem/logger.py
code served (permissive licence) · get_code("e6d5c8ead04a113a")
odestep Ran sisl/CEEM/ceem/odesolver.py
code served (permissive licence) · get_code("acefc50eb4bbfcac")
filter_ Not yet run sisl/CEEM/ceem/nested.py
code served (permissive licence) · get_code("333bf105fcfc9f91")
rk4_alt_step_func Not yet run sisl/CEEM/ceem/odesolver.py
code served (permissive licence) · get_code("5a92e97beb204564")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

System identification is a key step for modelbased control, estimator design, and output prediction. This work considers the offline identification of partially observed nonlinear systems. We empirically show that the certainty-equivalent approximation to expectation-maximization can be a reliable and scalable approach for highdimensional deterministic systems, which are common in robotics. We formulate certaintyequivalent expectation-maximization as block coordinate-ascent, and provide an efficient implementation. The algorithm is tested on a simulated system of coupled Lorenz attractors, demonstrating its ability to identify high-dimensional systems that can be intractable for particle-based approaches. Our approach is also used to identify the dynamics of an aerobatic helicopter. By augmenting the state with unobserved fluid states, a model is learned that predicts the acceleration of the helicopter better than state-of-the-art approaches. The codebase for this work is available at https://github.com/sisl/CEEM.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2006.11615")
get_code_for_paper("2006.11615")
have("2006.11615")

Connect an agent — have() is free.