PyBEAM: A Bayesian approach to parameter inference for a wide class of binary evidence accumulation models

Area of Science:

• Cognitive Science
• Computational Neuroscience
• Psychology

Background:

• Decision-making theories often use evidence accumulation models (EAMs).
• The Diffusion Decision Model (DDM) is a popular EAM but has limitations.
• Complex EAMs are computationally challenging, limiting their application.

Purpose of the Study:

• Introduce PyBEAM, a Python package for Bayesian Evidence Accumulation Models.
• Develop a general probabilistic framework for binary decision models.
• Address computational challenges in applying Bayesian methods to complex EAMs.

Main Methods:

• Developed a general probabilistic framework for choice and response time distributions in binary decision models.
• Integrated computationally optimized modeling with PyMC for Bayesian parameter estimation.
• Demonstrated parameter recovery for various EAMs.

Main Results:

• PyBEAM expands the range of EAMs applicable to Bayesian analysis.
• The package significantly reduces computational time for EAM analysis.
• PyBEAM lowers the barrier to entry for researchers using complex EAMs.

Conclusions:

• PyBEAM offers a powerful and accessible tool for advancing decision-making research.
• The package facilitates the application of complex EAMs to real-world data.
• This work enhances the computational and practical aspects of EAM analysis.