1 code implementation • 19 Mar 2024 • Aswin Paul, Takuya Isomura, Adeel Razi
Given the rapid advancement of artificial intelligence, understanding the foundations of intelligent behaviour is increasingly important.
no code implementations • 6 Dec 2023 • Karl J. Friston, Tommaso Salvatori, Takuya Isomura, Alexander Tschantz, Alex Kiefer, Tim Verbelen, Magnus Koudahl, Aswin Paul, Thomas Parr, Adeel Razi, Brett Kagan, Christopher L. Buckley, Maxwell J. D. Ramstead
First, we simulate the aforementioned in vitro experiments, in which neuronal cultures spontaneously learn to play Pong, by implementing nested, free energy minimising processes.
no code implementations • 16 Nov 2023 • Takuya Isomura
This work shows that the Hamiltonian of generic dynamical systems constitutes a class of generative models, thus rendering their Helmholtz energy naturally equivalent to variational free energy under the identified generative model.
no code implementations • 20 Nov 2021 • Manuel Baltieri, Takuya Isomura
In this work, we present a straightforward derivation of Kalman filters consistent with active inference via a variational treatment of free energy minimisation in terms of gradient descent.
no code implementations • 15 Nov 2021 • Takuya Isomura
Mathematical analyses demonstrate that a combination of the gradient descent algorithm and the selection and crossover algorithm--with a biased crossover weight--maximises the search efficiency.
1 code implementation • 1 Mar 2020 • Takuya Isomura, Taro Toyoizumi
Generalization of time series prediction remains an important open issue in machine learning, wherein earlier methods have either large generalization error or local minima.
1 code implementation • 2 Aug 2018 • Takuya Isomura, Taro Toyoizumi
This work theoretically validates that a cascade of linear PCA and ICA can solve a nonlinear BSS problem accurately -- when the sensory inputs are generated from hidden sources via nonlinear mappings with sufficient dimensionality.