AIExplainer

What is a dimension reduction?

Decreasing the number of dimensions used to represent a particular feature in a feature vector, typically by converting to an embedding vector.

Decreasing the number of dimensions used to represent a particular feature in a feature vector, typically by converting to an embedding vector.

Practitioners refer to dimension reduction when building, training, or evaluating machine learning systems. It appears in research papers, product documentation, and technical discussions about AI capabilities and limitations.