Content deleted Content added
m Cinder painter moved page Product quantization to Draft:Product quantization: Not ready for mainspace, incubate in draftspace. Reason/s: more sources needed |
AFC draft |
||
Line 1:
{{Draft article}}
'''Product quantization''' ('''PQ''') is a technique that decomposes high-dimensional [[Vector spaces | vector spaces]] into a [[Cartesian product]] of [[Dimension_(vector_space) | low-dimensional]] [[Linear_subspace | subspaces]], with each subspace quantized independently. This approach represents each vector by a compact code, enabling efficient distance estimation while significantly reducing memory usage. <ref>{{Cite web |title=Product Quantization for Nearest Neighbor Search |url=https://inria.hal.science/inria-00514462v2/document |access-date=2025-03-21 |website=hal.science }}</ref>
It is commonly used in approximate [[nearest neighbor search]], like the [[Hierarchical navigable small world]] data structure.
{{Drafts moved from mainspace|date=March 2025}}
| |||