Spherical Mixture Integration for Latent Embedding Alignment across Multi-Source Feature Spaces
Yuming Zhang, Congyuan Duan, Dong Xia, Doudou Zhou, Tianxi Cai · 2026
Read Spherical Mixture Integration for Latent Embedding Alignment across Multi-Source Feature Spaces by Yuming Zhang, Congyuan Duan, Dong Xia, Doudou Zhou, Tianxi Cai (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-08637.txt.
Description
Yuming Zhang, Congyuan Duan, Dong Xia, Doudou Zhou, and Tianxi Cai align patient record codes across hospitals that label the same condition differently. Pooling records from many institutions strengthens medical models, but local codes are granular and unmatched, splitting one concept into many fragments. Their SMILE method treats shared concept summaries as privacy safe embeddings and uses sparse known links as weak guidance.
Same meaning codes are modeled as clusters on a sphere, yielding one unified picture per clinical concept. The estimator comes with error bounds and reliable cluster recovery, and tests on simulations plus real multi hospital data show cleaner alignment. Health data research gains a principled way to merge records without sharing raw data.
Catalog record
| CREATOR | Yuming Zhang, Congyuan Duan, Dong Xia, Doudou Zhou, Tianxi Cai |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.08637