Most Efficient Landmark Subsets for Species and Sex Reclassification

A 9.5 KB Excel file presents results from a random search for the most efficient subsets of morphological landmarks. Author Jean-Pierre Dujardin used K-means to reclassify groups of males and females across multiple species, including Anopheles and Phlebotomus mosquitoes. The data compares classification accuracy using all landmarks versus the smallest optimal subset.

Use Cases

• Feature selection for morphometric classification based on landmark efficiency scores.
• Benchmarking K-means clustering performance on biological shape data.
• Comparing reclassification accuracy between sexes and species groups mentioned in the description.
• Optimizing landmark sets for efficient species identification in entomology.

Strengths

• Provides specific accuracy and adjusted accuracy scores for landmark subsets.
• Includes results for multiple species groups, such as Anopheles and Phlebotomus.
• Distinguishes between male and female classification tasks.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The 9.5 KB size indicates a very small, likely summary-level dataset.

Provenance

Source

Jean-Pierre Dujardin via figshare.

Collection Method

Likely derived from a random search and K-means clustering analysis on morphological landmark data.

Freshness

Last updated 2026-06-02 17:41:48; freshness should be verified.