liveBook · Manning

Quantization error

As explained in Build AI Drug Discovery Pipelines

Quantization error is the average distance between input data points and their best matching units in a self-organizing map.

For a self-organizing map (SOM), quantization error is the average distance between each data point and its best-matching unit (BMU). Lower values mean points are closer to their BMUs, indicating better representation. It is evaluated alongside topographic error; topographic error, not quantization error, is used to select the best model. section 7.1.2

Molecular fingerprints

SOM evaluation uses 2,048-dimensional Morgan fingerprints for 2,062 compounds. section 7.1.2

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Defines quantization error as average data-to-BMU distance and interprets lower values as better representation.
Examples
Evaluates maps using 2,048-dimensional Morgan fingerprints for 2,062 compounds.
What you can do after
Evaluate SOM representation quality from quantization error while selecting models with topographic error.
How it connects
Connects self-organizing maps with best-matching units, topographic error, and Morgan fingerprints.

Also explained in

Compare how 3 books explain quantization error →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build AI Drug Discovery Pipelines. The section links point to the parts of the book it draws on.