Krzysztof Siminski
Abstract:
The paper presents a memetic fuzzy inference system based on Big Bang Big Crunch (evolutionary optimisation) and gradient descent (local search) techniques. Tuning parameters of the fuzzy system with evolutionary optimisation failed to be successful, but application of both evolutionary and local optimisation achieved lower error rates than reference system (that uses only gradient descent optimisation). The results of experiments have been statistically verified.
Reference:
Krzysztof Siminski, Imputation of Missing Values by Inversion of Fuzzy Neuro-System, [chapter in] Man--Machine Interactions 4, Springer International Publishing, 2016, pp. 573-582.
Bibtex Entry:
@incollection{id:Siminski2016Imputation,
title={Imputation of Missing Values by Inversion of Fuzzy Neuro-System},
author={Siminski, Krzysztof},
booktitle={Man--Machine Interactions 4},
pages={573--582},
year={2016},
publisher={Springer International Publishing},
abstract = {The paper presents a memetic fuzzy inference system based
on Big Bang Big Crunch (evolutionary optimisation) and gradient descent
(local search) techniques. Tuning parameters of the fuzzy system
with evolutionary optimisation failed to be successful, but application
of both evolutionary and local optimisation achieved lower error rates
than reference system (that uses only gradient descent optimisation).
The results of experiments have been statistically verified.},
}