Research map / Time series / Causal and generative time series
Fast and Consistent Learning of Hidden Markov Models by Incorporating Non-Consecutive Correlations
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| include non-consecutive lags (up to 10) in second-order method of moments | second-order MoM for HMM transition matrix · synthetic X=Y=3 known-sensor HMM, N=10^5 | error of estimated P relative to tau=1: ~25% improvement | Helped |
| increase number of lags beyond roughly the chain's time constant (ablation) | multi-lag MoM · synthetic X=Y=3 HMM | error of estimated P | No clear effect |
| include non-consecutive lags in MoM instead of EM-like iterative ML | HMM parameter estimation · synthetic X=Y=3 HMM | run-time: almost two orders of magnitude faster than EM at 10^5 samples | Helped |
| include non-consecutive lags (tau=10) in second-order MoM | two-regime HMM on AAPL weekly log-returns · AAPL financial regimes | filter posterior resemblance to EM | Helped |
Closest papers by meaning
- Learning Hidden Markov Models When the Locations of Missing Observations are Unknown (ICML 2023 · similarity 0.87)
- Learning HMMs with Nonparametric Emissions via Spectral Decompositions of Continuous Matrices (NeurIPS 2016 · similarity 0.86)
- Scaling Factorial Hidden Markov Models: Stochastic Variational Inference without Messages (NeurIPS 2016 · similarity 0.86)
- The Segmented iHMM: A Simple, Efficient Hierarchical Infinite HMM (ICML 2016 · similarity 0.86)
- Learning Hidden Markov Models from Pairwise Co-occurrences with Application to Topic Modeling (ICML 2018 · similarity 0.86)
- Pre-trained Large Language Models Learn to Predict Hidden Markov Models In-context (NeurIPS 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.