CHARDA: Causal Hybrid Automata Recovery via Dynamic Analysis

Authors: Adam Summerville, Joseph Osborn, Michael Mateas

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We propose and evaluate a new technique for learning hybrid automata automatically by observing the runtime behavior of a dynamical system. In our experiments on a complex videogame character, CHARDA successfully discovers a reasonable over-approximation of the character s true behaviors. Our results also compare favorably against recent work in automatically learning probabilistic timed automata in an aircraft domain: CHARDA exactly learns the modes of these simpler automata. 5 Evaluation To evaluate our work we considered two domains: Aircraft Dynamics Modeling and Mario s Jump Dynamics from SMB.
Researcher Affiliation Academia Adam Summerville, Joseph Osborn, Michael Mateas University of California, Santa Cruz {asummerv, jcosborn} @ucsc.edu, michaelm@soe.ucsc.edu
Pseudocode No No structured pseudocode or algorithm blocks were found in the paper. The methodology is described in narrative text and mathematical equations.
Open Source Code No The paper does not provide any concrete access to source code for the methodology described, nor does it explicitly state that the code is released or available.
Open Datasets No The paper mentions 'Lawnmower data' and 'Mario trace' for its experiments. It cites Santana et al. [2015] for comparison in the aircraft domain and jdaster64 [2012] for a manually-defined Mario HA, but does not provide concrete access information (link, DOI, specific repository, or clear statement of public availability with author/year for the *specific data used*) for the datasets CHARDA was run on.
Dataset Splits No The paper does not provide specific dataset split information (e.g., exact percentages, sample counts, or detailed splitting methodology for training, validation, and test sets) needed for reproduction. It mentions '32 trials' for the aircraft domain and discarding 'the best and worst runs' but no explicit data splits.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types, or memory amounts) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details, such as library names with version numbers, needed to replicate the experiment.
Experiment Setup Yes The paper specifies using either the Bayesian Information Criterion (BIC) or Minimum Description Length (MDL) as penalty functions for model complexity, with their respective formulas. It also details the cost function C[i, j, m] = -log(L(T[i, j, m]|d[i : j])) + pen(m, d[i : j]) and mentions that 'for this work our set of models are all multivariate regressions'.