You are reviewing model benchmarks in Azure AI Foundry. You must use an embedding model that can assess rank-order relevance based on cosine similarity. You need to select the applicable embedding mod

The correct answer is B. Spearman correlation. Spearman correlation based on cosine similarity is calculated by first computing the cosine similarity between variables, then ranking these scores and using the ranks to compute the Spearman correlation. https://learn.microsoft.com/en-us/azure/ai-foundry/concepts/model-benchmark

Explore data and run experiments

Question

You are reviewing model benchmarks in Azure AI Foundry. You must use an embedding model that can assess rank-order relevance based on cosine similarity. You need to select the applicable embedding model. Which model metric should you focus on?

Options

AF1 score
BSpearman correlation
CV measure
DMean average precision

How the community answered

(27 responses)

A
11% (3)
B
78% (21)
C
7% (2)
D
4% (1)

Explanation

Spearman correlation based on cosine similarity is calculated by first computing the cosine similarity between variables, then ranking these scores and using the ranks to compute the Spearman correlation. https://learn.microsoft.com/en-us/azure/ai-foundry/concepts/model-benchmarks

Topics

#Embedding models#Model evaluation#Ranking metrics#Spearman correlation

Community Discussion

No community discussion yet for this question.

Full DP-100 Practice