PMID- 33286336
OWN - NLM
STAT- PubMed-not-MEDLINE
LR  - 20201210
IS  - 1099-4300 (Electronic)
IS  - 1099-4300 (Linking)
VI  - 22
IP  - 5
DP  - 2020 May 18
TI  - Goal-Directed Planning for Habituated Agents by Active Inference Using a 
      Variational Recurrent Neural Network.
LID - 10.3390/e22050564 [doi]
LID - 564
AB  - It is crucial to ask how agents can achieve goals by generating action plans 
      using only partial models of the world acquired through habituated sensory-motor 
      experiences. Although many existing robotics studies use a forward model 
      framework, there are generalization issues with high degrees of freedom. The 
      current study shows that the predictive coding (PC) and active inference (AIF) 
      frameworks, which employ a generative model, can develop better generalization by 
      learning a prior distribution in a low dimensional latent state space 
      representing probabilistic structures extracted from well habituated 
      sensory-motor trajectories. In our proposed model, learning is carried out by 
      inferring optimal latent variables as well as synaptic weights for maximizing the 
      evidence lower bound, while goal-directed planning is accomplished by inferring 
      latent variables for maximizing the estimated lower bound. Our proposed model was 
      evaluated with both simple and complex robotic tasks in simulation, which 
      demonstrated sufficient generalization in learning with limited training data by 
      setting an intermediate value for a regularization coefficient. Furthermore, 
      comparative simulation results show that the proposed model outperforms a 
      conventional forward model in goal-directed planning, due to the learned prior 
      confining the search of motor plans within the range of habituated trajectories.
FAU - Matsumoto, Takazumi
AU  - Matsumoto T
AUID- ORCID: 0000-0003-1388-2906
AD  - Okinawa Institute of Science and Technology, Okinawa 904-0495, Japan.
FAU - Tani, Jun
AU  - Tani J
AUID- ORCID: 0000-0002-9131-9206
AD  - Okinawa Institute of Science and Technology, Okinawa 904-0495, Japan.
LA  - eng
PT  - Journal Article
DEP - 20200518
PL  - Switzerland
TA  - Entropy (Basel)
JT  - Entropy (Basel, Switzerland)
JID - 101243874
PMC - PMC7517093
OTO - NOTNLM
OT  - active inference
OT  - goal directed planning
OT  - predictive coding
OT  - recurrent neural network
OT  - variational bayes
COIS- The authors declare no conflict of interest.
EDAT- 2020/12/09 06:00
MHDA- 2020/12/09 06:01
PMCR- 2020/05/18
CRDT- 2020/12/08 01:04
PHST- 2020/04/20 00:00 [received]
PHST- 2020/05/12 00:00 [revised]
PHST- 2020/05/15 00:00 [accepted]
PHST- 2020/12/08 01:04 [entrez]
PHST- 2020/12/09 06:00 [pubmed]
PHST- 2020/12/09 06:01 [medline]
PHST- 2020/05/18 00:00 [pmc-release]
AID - e22050564 [pii]
AID - entropy-22-00564 [pii]
AID - 10.3390/e22050564 [doi]
PST - epublish
SO  - Entropy (Basel). 2020 May 18;22(5):564. doi: 10.3390/e22050564.