Research map / Probabilistic and scientific modeling / Molecules and proteins
Learning Hidden Markov Models from Pairwise Co-occurrences with Application to Topic Modeling
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 |
|---|---|---|---|
| Use HTMM to model word order with Markovian topic transitions instead of a bag-of-words model | Reuters21578 topic modeling · Reuters21578 | topic coherence | Helped |
| Use HTMM instead of bag-of-words topic models | Reuters21578 topic modeling · Reuters21578 | perplexity | Helped |
| Use pairwise co-occurrence factorization with the determinant-based identifiability criterion | Synthetic HMM learning with N=100 and K=20 · Synthetic HMM data | total variation distance | Helped |
| Use pairwise co-occurrence factorization with the determinant-based identifiability criterion instead of plain NMF | Synthetic HMM learning · Synthetic HMM data | parameter estimation error | Helped |
| Use pairwise co-occurrence factorization instead of tensor CPD | Synthetic HMM learning · Synthetic HMM data | parameter estimation error | Helped |
| Use pairwise co-occurrence factorization instead of CPD-initialized EM | Synthetic HMM learning · Synthetic HMM data | parameter estimation error | Helped |
| Use pairwise co-occurrence factorization instead of Baum-Welch | Synthetic HMM learning with N=16 and K=4 · Synthetic HMM data | parameter estimation error | Helped |
Closest papers by meaning
- Learning Hidden Markov Models When the Locations of Missing Observations are Unknown (ICML 2023 · similarity 0.86)
- Fast and Consistent Learning of Hidden Markov Models by Incorporating Non-Consecutive Correlations (ICML 2020 · similarity 0.86)
- Learning HMMs with Nonparametric Emissions via Spectral Decompositions of Continuous Matrices (NeurIPS 2016 · similarity 0.85)
- An Infinite Hidden Markov Model With Similarity-Biased Transitions (ICML 2017 · similarity 0.84)
- FlowHMM: Flow-based continuous hidden Markov models (NeurIPS 2022 · similarity 0.84)
- Pre-trained Large Language Models Learn to Predict Hidden Markov Models In-context (NeurIPS 2025 · similarity 0.81)
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.