Parallel Sorting Algorithm Performance Profiles for Machine Learning

Comprising Caliper performance profiles for analyzing parallel sorting algorithm performance, used in research on algorithm classification. It was created by Michael McKinsey for the study 'Parallel sorting algorithm classification: is manual instrumentation necessary?'.

Use Cases

• Train a machine learning classifier to predict the parallel sorting algorithm class using performance profile data.
• Analyze performance characteristics from Caliper profiles to compare the efficiency of different parallel sorting implementations.
• Use the performance data to investigate correlations between specific algorithm features and execution metrics for classification research.

Strengths

• Data is specifically designed for machine learning classification of parallel sorting algorithms.
• Profiles are generated using the established Caliper performance measurement framework.

Limitations

• The dataset size, row count, and specific features are unknown, limiting initial assessment of scope.
• Data format requires specialized tools like Thicket for analysis, creating a barrier to entry.

Provenance

Source

Texas Data Repository Harvested Dataverse

Collection Method

Performance profiles generated using Caliper for parallel sorting algorithm analysis.

Freshness

Last updated on 2026-02-16.