Table 1.
Applications of active inference for Markov decision processes.
| Application | Comment | References |
|---|---|---|
| Decision making under uncertainty | Initial formulation of active inference for Markov decision processes and sequential policy optimisation | (Friston et al., 2012b) |
| Optimal control (the mountain car problem) | Illustration of risk sensitive or KL control in an engineering benchmark | (Friston et al., 2012a) |
| Evidence accumulation: Urns task | Demonstration of how beliefs states are absorbed into a generative model | (FitzGerald et al., 2015b,c) |
| Addiction | Application to psychopathology | (Schwartenbeck et al., 2015c) |
| Dopaminergic responses | Associating dopamine with the encoding of (expected) precision provides a plausible account of dopaminergic discharges | (Friston et al., 2014 ; FitzGerald et al., 2015a) |
| Computational fMRI | Using Bayes optimal precision to predict activity in dopaminergic areas | (Schwartenbeck et al., 2015a) |
| Choice preferences and epistemics | Empirical testing of the hypothesis that people prefer to keep options open | (Schwartenbeck et al., 2015b) |
| Behavioural economics and trust games | Examining the effects of prior beliefs about self and others | (Moutoussis et al., 2014) |
| Foraging and two step mazes | Formulation of epistemic and pragmatic value in terms of expected free energy | (Friston et al., 2015) |
| Habit learning, reversal learning and devaluation | Learning as minimising variational free energy with respect to model parameters – and action selection as Bayesian model averaging | (FitzGerald et al., 2014; Friston et al., 2016) |
| Saccadic searches and scene construction | Mean field approximation for multifactorial hidden states, enabling high dimensional beliefs and outcomes: c.f., functional segregation | (Friston and Buzsaki, 2016; Mirza et al., 2016) |
| Electrophysiological responses: place-cell activity, omission related responses, mismatch negativity, P300, phase-procession, theta-gamma coupling | Simulating neuronal processing with a gradient descent on variational free energy; c.f., dynamic Bayesian belief propagation based on marginal free energy | In press |
| Structure learning, sleep and insight | Inclusion of parameters into expected free energy to enable structure learning via Bayesian model reduction | Under review |
| Narrative construction and reading | Hierarchical generalisation of generative model with deep temporal structure | Current paper |