Integration of Node Embeddings for Multiple Versions of A Network
| Main Author: | |
|---|---|
| Language: | English |
| Published: |
Case Western Reserve University School of Graduate Studies / OhioLINK
2020
|
| Subjects: | |
| Online Access: | http://rave.ohiolink.edu/etdc/view?acc_num=case1595435155975104 |
| id |
ndltd-OhioLink-oai-etd.ohiolink.edu-case1595435155975104 |
|---|---|
| record_format |
oai_dc |
| spelling |
ndltd-OhioLink-oai-etd.ohiolink.edu-case15954351559751042021-08-15T09:57:36Z Integration of Node Embeddings for Multiple Versions of A Network Li, Mengzhen Computer Science Machine learning applications on large-scale network-structured data commonly encode network information in the form of node embeddings. The general principle of existing algorithms for computing network embeddings is to map the nodes into a low-dimensional space such that the nodes that are “similar" with respect to network topology are also close to each other in the embedding space. Many real-world networks that are used in machine learning have multiple versions that come from different sources, are stored in different databases, or belong to different parties. Due to efficiency or privacy concerns, it may be desirable to compute consensus embeddings for the superposed network directly from the node embeddings of individual versions, without explicitly constructing the superposed network. We consider multiple approaches to compute embeddings for the superposed network from the embeddings of individual versions. To systematically assess the quality of the resulting consensus embeddings, we define the notion of fidelity. We then test the performance of consensus embeddings on link prediction. Our results show that predictions obtained with consensus embeddings are almost as accurate as those that are obtained with embeddings computed using the superposed network. 2020-09-07 English text Case Western Reserve University School of Graduate Studies / OhioLINK http://rave.ohiolink.edu/etdc/view?acc_num=case1595435155975104 http://rave.ohiolink.edu/etdc/view?acc_num=case1595435155975104 unrestricted This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws. |
| collection |
NDLTD |
| language |
English |
| sources |
NDLTD |
| topic |
Computer Science |
| spellingShingle |
Computer Science Li, Mengzhen Integration of Node Embeddings for Multiple Versions of A Network |
| author |
Li, Mengzhen |
| author_facet |
Li, Mengzhen |
| author_sort |
Li, Mengzhen |
| title |
Integration of Node Embeddings for Multiple Versions of A Network |
| title_short |
Integration of Node Embeddings for Multiple Versions of A Network |
| title_full |
Integration of Node Embeddings for Multiple Versions of A Network |
| title_fullStr |
Integration of Node Embeddings for Multiple Versions of A Network |
| title_full_unstemmed |
Integration of Node Embeddings for Multiple Versions of A Network |
| title_sort |
integration of node embeddings for multiple versions of a network |
| publisher |
Case Western Reserve University School of Graduate Studies / OhioLINK |
| publishDate |
2020 |
| url |
http://rave.ohiolink.edu/etdc/view?acc_num=case1595435155975104 |
| work_keys_str_mv |
AT limengzhen integrationofnodeembeddingsformultipleversionsofanetwork |
| _version_ |
1719460024240046080 |