Estimating Latent-Variable Graphical Models using Moments and Likelihoods

Authors: Arun Tejasvi Chaganty, Percy Liang

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

Reproducibility Variable Result LLM Response
Research Type Experimental Figure 4. Comparison of parameter estimation error (ˆθ − θ 2) versus error in moments (ϵ) for a hidden Markov model with k = 2 hidden and d = 5 observed values. Empirical moments c M123 were generated by adding Gaussian noise, N(0, ϵI), to expected moments M123. Results are averaged over 400 trials.
Researcher Affiliation Academia Arun Tejasvi Chaganty CHAGANTY@CS.STANFORD.EDU Percy Liang PLIANG@CS.STANFORD.EDU Stanford University, Stanford, CA, USA
Pseudocode Yes Algorithm 1 GETMARGINALS (pseudoinverse)
Open Source Code No The paper does not contain any statement about releasing source code for the methodology described, nor does it provide any links to a code repository.
Open Datasets No Empirical moments c M123 were generated by adding Gaussian noise, N(0, ϵI), to expected moments M123.
Dataset Splits No The paper discusses the comparison of statistical efficiency between estimators using generated empirical moments over 'n data points' (Section 4.3), but it does not specify train/validation/test dataset splits for reproduction.
Hardware Specification No The paper mentions numerical experiments were performed and results averaged over 400 trials, but it does not provide any specific hardware details such as GPU/CPU models or memory specifications.
Software Dependencies No The paper refers to optimization procedures like 'EM or LBFGS' but does not specify any software libraries, packages, or their version numbers that were used for implementation or experimentation.
Experiment Setup Yes Comparison of parameter estimation error (ˆθ − θ 2) versus error in moments (ϵ) for a hidden Markov model with k = 2 hidden and d = 5 observed values. Empirical moments c M123 were generated by adding Gaussian noise, N(0, ϵI), to expected moments M123.