Bridging the Gap Between Theory and Practice of Approximate Bayesian Inference

A paper discussing the computational complexity of Bayesian models in cognitive science. Authored by Johan Kwisthout of Leiden University, it examines three candidate notions of approximation to make intractable Bayesian computations tractable for modeling human cognition.

Use Cases

• Developing computationally plausible cognitive models based on the discussed approximation methods.
• Analyzing the trade-offs between model accuracy and computational tractability in Bayesian inference.
• Parameterizing computational complexity analyses to yield tractable model variants.
• Informing the design of experiments in cognitive science that test approximate Bayesian models.

Strengths

• Authored by a researcher from Leiden University, suggesting academic rigor.
• Focuses on a specific theoretical challenge in computational cognitive science.
• Discusses three distinct candidate notions of 'approximation' from the literature.

Limitations

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

Provenance

Source

paperswithcode