Deterministic Anytime Inference for Stochastic Continuous-Time Markov Processes

Authors: E. Busra Celikkaya, Christian Shelton

ICML 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We implemented our method, TTOP (Tree of Time Ordered Products), as part of the CTBN-RLE code base (Shelton et al., 2010), and it will be included in the next version. We evaluated our method on a synthetic network of Ising model dynamics.
Researcher Affiliation Academia E. Busra Celikkaya CELIKKAE@CS.UCR.EDU University of California, Riverside Christian R. Shelton CSHELTON@CS.UCR.EDU University of California, Riverside
Pseudocode Yes Algorithm 1 TTOP Filter
Open Source Code No We implemented our method, TTOP (Tree of Time Ordered Products), as part of the CTBN-RLE code base (Shelton et al., 2010), and it will be included in the next version.
Open Datasets No We evaluated our method on a synthetic network of Ising model dynamics. The Ising model was chosen so that we could compute the true answer in a reasonable time and scale the problem size. Using this model, we generated a directed toroid network structure with cycles following (El-Hay et al., 2010).
Dataset Splits No The paper uses a synthetic network and does not specify explicit training, validation, or test dataset splits.
Hardware Specification No No specific hardware details (like GPU/CPU models or types) used for running experiments were mentioned.
Software Dependencies No The paper mentions implementing the method as part of the 'CTBN-RLE code base' but does not list specific software dependencies with version numbers.
Experiment Setup Yes For TTOP, we set the number of splits for the quadrature (see Equation 5) to 10 because it produces a good computation time versus error performance. ... For Aux Gibbs, we vary the sample size between 50 and 5000, and set the burn-in period to be 10% of this value. For IS, the sample size varies between 500 to 50000.