Formal frameworks that use [[Mathematics|mathematical]] or [[Algorithms & Data Structures|algorithmic]] representations to simulate mental processes and neural dynamics. They provide testable predictions, allow comparison with behavioral and neural data, and bridge [[Psychology]], [[Neuroscience]], and [[AI]].
- Core Approaches
- Symbolic models
- rule-based, logical structures (e.g., production systems)
- Connectionist models
- neural networks that learn through weighted associations
- Bayesian models
- probabilistic inference and [[decision-making]] under uncertainty
- Reinforcement learning
- value-based choice and reward-driven learning
- Hybrid models
- integrate symbolic reasoning with network-level learning
- Applications
- [[Memory]]
- models of storage, retrieval, and forgetting curves
- [[Decision-Making]]
- reinforcement learning, drift-diffusion models
- [[Attention]]
- resource allocation, signal-to-noise dynamics
- [[Language Processing]]
- parsing, semantic networks, deep learning models
- [[Emotion]]
- affective computing, appraisal models
- Perspectives
- [[Cognitive Psychology]]
- explains behavior via abstract rules or processes
- [[Cognitive Neuroscience]]
- links algorithms to brain activity (fMRI/EEG data fitting)
- [[Artificial Intelligence]]
- builds intelligent systems inspired by cognition
- Subtopics
- [[Connectionist Models]]
- [[Bayesian Models]]
- [[Reinforcement Learning]]
- [[Drift-Diffusion Model]]
- Methods
- Simulation experiments
- test predictions against behavioral data
- Neural network architectures
- layers, weights, activation functions
- Probabilistic inference
- likelihood estimation, uncertainty quantification