ML Anthology

Learnable Similarity Functions and Their Applications to Clustering and Record Linkage

Mikhail Bilenko
AAAI 2004 pp. 981-982
/aaai/2004/bilenko2004aaai-learnable/

Abstract

Many problems in machine learning and data mining depend on distance estimates between observations, e.g., instance-based classification, clustering, information re-trieval, and record linkage in databases. However, the ap-propriate notion of similarity can vary depending on the par-ticular domain, dataset, or task at hand. Consequently, a large number of functions that compute similarity between objects have been developed for different data types, vary-ing greatly in their expressiveness, mathematical properties, and assumptions (Gusfield 1997; Duda, Hart, & Stork 2001). Additionally, there exists a substantial body of research on feature space transformations that attempt to provide a more salient representation of data than the original feature space,

PDF AAAI Semantic Scholar

Cite

Text

Bilenko. "Learnable Similarity Functions and Their Applications to Clustering and Record Linkage." AAAI Conference on Artificial Intelligence, 2004.

Markdown

[Bilenko. "Learnable Similarity Functions and Their Applications to Clustering and Record Linkage." AAAI Conference on Artificial Intelligence, 2004.](https://mlanthology.org/aaai/2004/bilenko2004aaai-learnable/)

BibTeX

@inproceedings{bilenko2004aaai-learnable,
  title     = {{Learnable Similarity Functions and Their Applications to Clustering and Record Linkage}},
  author    = {Bilenko, Mikhail},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2004},
  pages     = {981-982},
  url       = {https://mlanthology.org/aaai/2004/bilenko2004aaai-learnable/}
}