Learning Stable Stochastic Nonlinear Dynamical Systems
Authors: Jonas Umlauft, Sandra Hirche
ICML 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We validate our approach, labeled Le SSS (for Learning Stable Stochastic Systems), using synthetic and human motion data and the simulation of a chemical reactor. |
| Researcher Affiliation | Academia | 1Chair of Information-oriented Control, Technical University of Munich, Munich, Germany. |
| Pseudocode | No | The paper describes its learning framework in three major steps (Section 4) but does not provide structured pseudocode or an algorithm block. |
| Open Source Code | No | The paper does not provide any concrete access information (e.g., a specific repository link, explicit code release statement, or code in supplementary materials) for the open-source code of the methodology described. |
| Open Datasets | Yes | For the next simulation, we use the data set for letter-shaped motions provided by Khansari-Zadeh & Billard (2011). For the last validation, we utilize simulated data from a simplified chemical reactor (Einarsson, 1998). |
| Dataset Splits | No | The paper describes the training data used for simulations but does not provide specific information about train/validation/test splits, such as percentages, sample counts, or cross-validation setup. |
| Hardware Specification | Yes | The computation times on a i5 CPU 2.30GHz, 2 Cores and 8GB RAM are given for Simulation 2 and 3 in Table 2. |
| Software Dependencies | No | The paper mentions using GMR and GPDM (Section 5) and refers to 'The code (based on Calinon (2009))' but does not provide specific version numbers for any software dependencies or libraries used in the experiments. |
| Experiment Setup | Yes | Here L = 3 was chosen for the number of mixtures in the GMR for the mapping θψ B : X ΘB. For our simulations, five mixtures are employed. The 2d = 4 closest data points are considered for fitting the training parameters of the Dirichlet distribution locally. perturbed with white noise with σ = 0.01 for both states. |