L. Ferariu - Profile on Academia.edu (original) (raw)

Lavinia Ferariu related author profile picture

Bogdan Burlacu related author profile picture

DR KUNTAL BARUA related author profile picture

Silvia Curteanu related author profile picture

Dr. Ashraf M M Aboshosha related author profile picture

Lâm Vũ Minh Hưng related author profile picture

Chintan Gajjar related author profile picture

Miguel Benitez related author profile picture

Jun S. Huang related author profile picture

Prof. Mary Opokua Ansong related author profile picture

Uploads

Papers by L. Ferariu

Research paper thumbnail of The design of hybrid neural networks by means of genetic programming

The Bulletin of Politechnical Institute of Iasi, Jan 1, 2010

This paper presents a new genetic programming based approach for the design of feed forward hybri... more This paper presents a new genetic programming based approach for the design of feed forward hybrid neural models. A special encryption of the neural model as a directed acyclic graph (DAG) is suggested, intended to exploit the modularity of the neural topology. This allows for a flexible development of partially interconnected neural structures, having heterogeneous layers with both Gaussian and perceptron neurons. Customized compatible genetic operators, aimed to work simultaneously on the neural architecture and parameters, provide an efficient exploration of the search space. Additionally, the algorithm makes use of a back-propagation procedure, employed as a local Lamarckian optimization, for a faster computation of model parameters. The performances of the suggested approach are illustrated on the identification of an industrial subsystem from the Sugar factory of Lublin, Poland.

Research paper thumbnail of The design of hybrid neural networks by means of genetic programming

The Bulletin of Politechnical Institute of Iasi, Jan 1, 2010

This paper presents a new genetic programming based approach for the design of feed forward hybri... more This paper presents a new genetic programming based approach for the design of feed forward hybrid neural models. A special encryption of the neural model as a directed acyclic graph (DAG) is suggested, intended to exploit the modularity of the neural topology. This allows for a flexible development of partially interconnected neural structures, having heterogeneous layers with both Gaussian and perceptron neurons. Customized compatible genetic operators, aimed to work simultaneously on the neural architecture and parameters, provide an efficient exploration of the search space. Additionally, the algorithm makes use of a back-propagation procedure, employed as a local Lamarckian optimization, for a faster computation of model parameters. The performances of the suggested approach are illustrated on the identification of an industrial subsystem from the Sugar factory of Lublin, Poland.

Log In