DiceOptim: Kriging-Based Optimization for Computer Experiments

DiceOptim is an R package implementing the Efficient Global Optimization (EGO) algorithm for computer experiments. It was authored by Victor Picheny and includes adaptations for problems with noise, parallel infill, and constraints as described in referenced papers from 2012 and 2013. The dataset likely contains simulation or experimental results used to demonstrate and benchmark these optimization methods.

Use Cases

• Benchmarking optimization algorithms based on the described EGO method.
• Optimizing simulation parameters for engineering design based on Kriging models.
• Handling optimization problems with noisy objective functions as referenced in the description.
• Implementing parallel infill strategies for computer experiments.
• Solving constrained optimization problems in a simulation-based context.

Strengths

• Implements a well-documented algorithm (EGO) from peer-reviewed publications.
• Includes specific adaptations for noise, parallelism, and constraints as cited.
• Authored by a named contributor (Victor Picheny) associated with the referenced research.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Victor Picheny

Collection Method

Likely a software package or code repository for statistical optimization.