Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Online Trans-dimensional von Mises-Fisher Mixture Models for User Profiles
Xiangju Qin, Pádraig Cunningham, Michael Salter-Townshend
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 TvMFMM (reversible jump, estimates H) instead of fixed-H BvMFMM | vMF mixture · Synthetic vMF (7 not well-separated components) | ARI: 0.612 -> 0.809 | Improved |
| use TvMFMM instead of BvMFMM | vMF mixture · Synthetic vMF (4 well-separated components) | ARI: 0.881 -> 0.925 | No clear effect |
| use TvMFMM instead of DP-GMM | vMF mixture · Synthetic vMF (5 well-separated components) | ARI: 0.599 -> 0.947 | Improved |
| use TvMFMM instead of k-means (k-means++) | clustering · Synthetic vMF (4 well-separated components) | ARI: 0.517 -> 0.925 | Improved |
| use TvMFMM instead of k-means (k-means++) | clustering · Synthetic vMF (5 well-separated components) | Homogeneity: 0.958 -> 0.947 | No clear effect |
| use common eigenvector matrix in split-merge moves (ablation) | TvMFMM · Synthetic vMF (4 well-separated components) | ARI: 0.931 -> 0.925 | No clear effect |
| use common eigenvector matrix in split-merge moves (ablation) | TvMFMM · Synthetic vMF (7 not well-separated components) | ARI: 0.792 -> 0.809 | No clear effect |
| use common eigenvectors in split-merge moves (ablation) | TvMFMM · Synthetic vMF (7 not well-separated components) | split-merge acceptance rate (%): 0.42 -> 0.69 | Improved |
| initialise OTvMFMM from previous quarter posterior instead of k-means++ | OTvMFMM · Wikipedia editors Quarter 15 | log-likelihood convergence | Improved |
| add temporal OTvMFMM over NMF/DTM | OTvMFMM · Wikipedia editors quarters 10-30 | NPMI coherence | Improved |
| use OTvMFMM instead of BvMFMM | OTvMFMM · Wikipedia editors quarters 10-30 | held-out log likelihood | Improved |
| use OTvMFMM instead of other models | OTvMFMM · Wikipedia editors Quarter 13 | learning time (s): 735.66 -> 53333.33 | Hurt |
Closest papers by meaning
- Consistent Model-based Clustering using the Quasi-Bernoulli Stick-breaking Process (JMLR 2023 · similarity 0.84)
- Online but Accurate Inference for Latent Variable Models with Local Gibbs Sampling (JMLR 2017 · similarity 0.83)
- CrossCat: A Fully Bayesian Nonparametric Method for Analyzing Heterogeneous, High Dimensional Data (JMLR 2016 · similarity 0.83)
- Global-Local Dirichlet Processes for Clustering Grouped Data in the Presence of Group-Specific Idiosyncratic Variables (ICML 2025 · similarity 0.82)
- Nemotron-CLIMB: Clustering-based Iterative Data Mixture Bootstrapping for Language Model Pre-training (NeurIPS 2025 · similarity 0.82)
- MF-LLM: Simulating Population Decision Dynamics via a Mean-Field Large Language Model Framework (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.
