Artificial Neural Maps

The study of artificial neural map formation
provides useful tools for biological and statistical
data modelling. "Artificial Neural Maps"
offers a systematic survey of extended neural map
concepts and architectures.
In the first part of this work, several new extended
neural map architectures are introduced. This kind
of modularity has been observed in the
organization of the human cortex, where sensory,
computational and motor maps are arranged in a
similar fashion. This modular approach
provides increased flexibility in customizing
architectures to specific problems. Example
applications include: system identification,
vibration control, gene expression analysis, inverse
kinematics of robots, biomedical signal
visualization and plant recognition.
In the second part of this work new neural network
based techniques are presented for performing multi-
sensor data fusion. The presented techniques are
applied in several data fusion applications where
heterogeneous sensor data are available.
The target group includes neural network
researchers, industrial scientists involved in data
mining and researchers in biosystems engineering.

Autorentext

Dimitrios Moshou, PhD: Studied Electrical Engineering at Democritus University of Thrace, Greece, Control and Information Technology at Manchester University, UK and Computational Intelligence (PhD) at K.U. Leuven, Belgium. He has performed research in biosystems engineering. He is an Assistant Professor at Aristotle University, Greece.

Klappentext

The study of artificial neural map formation provides useful tools for biological and statistical data modelling. "Artificial Neural Maps" offers a systematic survey of extended neural map concepts and architectures.In the first part of this work, several new extended neural map architectures are introduced. This kind of modularity has been observed in the organization of the human cortex, where sensory, computational and motor maps are arranged in a similar fashion. This modular approach provides increased flexibility in customizing architectures to specific problems. Example applications include: system identification, vibration control, gene expression analysis, inverse kinematics of robots, biomedical signal visualization and plant recognition.In the second part of this work new neural network based techniques are presented for performing multi-sensor data fusion. The presented techniques are applied in several data fusion applications where heterogeneous sensor data are available.The target group includes neural network researchers, industrial scientists involved in data mining and researchers in biosystems engineering.