Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Identification of Analytic Nonlinear Dynamical Systems with Non-asymptotic Guarantees
Authors: Negin Musavi, Ziyao Guo, Geir Dullerud, Yingying Li
NeurIPS 2024 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Lastly, we numerically compare our theoretical bounds with the empirical performance of LSE and SME on a pendulum example and a quadrotor example. Numerically, we test our theoretical results in pendulum and quadrotor systems. Simulations show that LSE and SME can indeed efficiently explore the system and converge to the true parameter under non-active exploration noises. |
| Researcher Affiliation | Academia | Negin Musavi EMAIL Ziyao Guo EMAIL Geir Dullerud EMAIL Yingying Li EMAIL Coordinated Science Laboratory University of Illinois Urbana-Champaign |
| Pseudocode | No | The paper does not contain any pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | Yes | Further details can be found in our source code.7 7https://github.com/Negin Musavi/real-analytic-nonlinear-sys-id |
| Open Datasets | No | Lastly, we numerically compare our theoretical bounds with the empirical performance of LSE and SME on a pendulum example and a quadrotor example. The ground truth for the unknown parameters for pendulum example in Example 1 is set to be m = 0.1 (kg), l = 0.5 (m)... |
| Dataset Splits | No | The paper does not specify training, validation, or test dataset splits. The experiments involve numerical simulations of dynamical systems rather than traditional machine learning dataset splits. |
| Hardware Specification | No | The paper mentions running 'Simulations' but does not provide specific details on the hardware used for these numerical experiments. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers needed to replicate the experiment. |
| Experiment Setup | Yes | The discretization time step in our numerical experiments is dt = 0.01 (s). The control input is a simple feedback controller ut = k αt +ηt. In Figures 1a and 1b we choose k = 2 and in Figures 2a, 2b and 3 we choose k = 0.1. The controller gains in our numerical experiments are chosen as: kpz = 0.75, kdz = 1.25, kpϕ = 0.03, kdϕ = 0.00875, kpθ = 0.03, kdθ = 0.00875, kpψ = 0.03, kdψ = 0.00875. For our experiments, we employ noise and disturbances drawn from uniform and truncated-Gaussian distributions. |