TY - GEN
T1 - Causal blankets
T2 - 1st International Workshop on Active Inference, IWAI 2020 held in conjunction with the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2020
AU - Rosas, Fernando E.
AU - Mediano, Pedro A.M.
AU - Biehl, Martin
AU - Chandaria, Shamil
AU - Polani, Daniel
N1 - Funding Information:
F.R. was supported by the Ad Astra Chandaria foundation. P.M. was funded by the Wellcome Trust (grant no. 210920/Z/18/Z). M.B. was supported by a grant from Tem-pleton World Charity Foundation, Inc. (TWCF). The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of TWCF.
Publisher Copyright:
© 2020, Springer Nature Switzerland AG.
This is a post-peer-review, pre-copyedit version of Rosas, F. E., Mediano, P. A. M., Biehl, M., Chandaria, S., & Polani, D. (2020). Causal blankets: Theory and algorithmic framework. In T. Verbelen, P. Lanillos, C. L. Buckley, & C. De Boom (Eds.), Active Inference - First International Workshop, IWAI 2020, Co-located with ECML/PKDD 2020, Proceedings (pp. 187-198). (Communications in Computer and Information Science; Vol. 1326). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-64919-7_19
PY - 2020/12/18
Y1 - 2020/12/18
N2 - We introduce a novel framework to identify perception-action loops (PALOs) directly from data based on the principles of computational mechanics. Our approach is based on the notion of causal blanket, which captures sensory and active variables as dynamical sufficient statistics—i.e. as the “differences that make a difference.” Furthermore, our theory provides a broadly applicable procedure to construct PALOs that requires neither a steady-state nor Markovian dynamics. Using our theory, we show that every bipartite stochastic process has a causal blanket, but the extent to which this leads to an effective PALO formulation varies depending on the integrated information of the bipartition.
AB - We introduce a novel framework to identify perception-action loops (PALOs) directly from data based on the principles of computational mechanics. Our approach is based on the notion of causal blanket, which captures sensory and active variables as dynamical sufficient statistics—i.e. as the “differences that make a difference.” Furthermore, our theory provides a broadly applicable procedure to construct PALOs that requires neither a steady-state nor Markovian dynamics. Using our theory, we show that every bipartite stochastic process has a causal blanket, but the extent to which this leads to an effective PALO formulation varies depending on the integrated information of the bipartition.
KW - Computational mechanics
KW - Integrated information
KW - Perception-action loops
KW - Stochastic processes
UR - http://www.scopus.com/inward/record.url?scp=85098286677&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-64919-7_19
DO - 10.1007/978-3-030-64919-7_19
M3 - Conference contribution
AN - SCOPUS:85098286677
SN - 9783030649180
T3 - Communications in Computer and Information Science
SP - 187
EP - 198
BT - Active Inference - First International Workshop, IWAI 2020, Co-located with ECML/PKDD 2020, Proceedings
A2 - Verbelen, Tim
A2 - Lanillos, Pablo
A2 - Buckley, Christopher L.
A2 - De Boom, Cedric
PB - Springer Nature
Y2 - 14 September 2020 through 14 September 2020
ER -