Kernel Recursive ABC: Point Estimation with Intractable Likelihood

Authors: Takafumi Kajihara, Motonobu Kanagawa, Keisuke Yamazaki, Kenji Fukumizu

ICML 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We have conducted a variety of numerical experiments, including parameter estimation for a real-world pedestrian flow simulator, and show that in most cases our method outperforms existing approaches.
Researcher Affiliation Collaboration 1NEC Corporation 2National Institute of Advanced Industrial Science and Technology 3Max Planck Institute for Intelligent Systems 4The Institute of Statistical Mathematics.
Pseudocode Yes Algorithm 1 Kernel Recursive ABC
Open Source Code No The paper mentions using 'publicly available code3' for comparison with a third-party method (Bayesian Optimization), but does not state that the code for their proposed Kernel Recursive ABC method is open-source or publicly available.
Open Datasets Yes Crowd Walk, a publicly available real-world simulator5 for the movements of pedestrians in a commercial district (Yamashita et al., 2010). Footnote 5: https://github.com/crest-cassia/CrowdWalk
Dataset Splits Yes That is, to evaluate one configuration of hyper-parameters, we first used 75% of the observed data for point estimation and then computed the discrepancy between the rest of the observed data and the ones simulated from point estimates
Hardware Specification No The paper does not provide specific details about the hardware used for running its experiments.
Software Dependencies No The paper mentions software like 'GPyOpt' (for Bayesian optimization) but does not provide specific version numbers for any software dependencies used in its experiments.
Experiment Setup Yes The bandwidth of a Gaussian kernel was selected from candidate values, each of which is the median (of pairwise distances) multiplied by logarithmically equally spaced values between 2 4 and 24 (Takeuchi et al., 2006, Sec. 5.1.1). Regularization constants for the proposed method and kernel ABC, as well as the soft threshold for K2-ABC, were selected from logarithmically spaced values between 10 4 and 1.