Biological Neural Networks: Hierarchical Concept of Brain Function

Thomas H. Jobe

This volume is a systematic presentation of the author's theory of neurocomputation. He has developed the "functional automatism" theory of neurocomputing to a high level of conceptual clarity. He shows how Kolmogorov's theorem (1957) can be applied to the new concept of a "spiking" dendritic tree to explain learning in reflex motor activity and how hierarchies of reflexes can control manipulation of external objects by an organism. The author has successfully applied the functional automatism concept of neurocomputing to motor reflex learning and control. His purpose is also to critique competing theories of neurocomputation. This book is intended for neurobiologists, neurologists, neurosurgeons and neuropsychiatrists. It provides an ingenious model of Parkinson's Disease that explains how neurosurgical interventions that damage tissue can actually improve function. Several excellent diagrams and illustrations are included that illustrate concepts of motor control by hierarchical reflex integration. Experiments that illustrate "functional automatisms" are well illustrated. The author has achieved a lucid and well-argued presentation of his model of neurocomputation. The strength of the model rests on the fact that it originally derives from biological control systems (vonHolst) and then makes use of important analogical extensions to achieve generality and complexity not found in more "mechanical" physics based models such as those of Hopfield and Tank. The author's approach thus stands as an important biological alternative to the dominant mechanistic trend in neural network theory.

Booknews

Presents a novel conceptual framework for neurobiology, achieved by applying control theory. It provides unifying principles for understanding the functional construction of the nervous system, and argues that all hierarchical levels of the nervous system are built according to the same functional principles, which underlie the highest forms of brain function. Within the framework, each network hierarchy is structurally and functionally organized so that a lower control system becomes the controlled object for a higher one, and each level of control processes a behavioral model of its controlled object. For neuroscientists, computer scientists and mathematicians, physicists, and clinicians interesting in deciphering how the brain works. Annotation c. by Book News, Inc., Portland, Or.

From The Critics

Reviewer: Thomas H. Jobe, MD(University of Illinois at Chicago College of Medicine)
Description: This volume is a systematic presentation of the author's theory of neurocomputation. He has developed the "functional automatism" theory of neurocomputing to a high level of conceptual clarity. He shows how Kolmogorov's theorem (1957) can be applied to the new concept of a "spiking" dendritic tree to explain learning in reflex motor activity and how hierarchies of reflexes can control manipulation of external objects by an organism.
Purpose: The author has successfully applied the functional automatism concept of neurocomputing to motor reflex learning and control. His purpose is also to critique competing theories of neurocomputation.
Audience: This book is intended for neurobiologists, neurologists, neurosurgeons and neuropsychiatrists. It provides an ingenious model of Parkinson's Disease that explains how neurosurgical interventions that damage tissue can actually improve function.
Features: Several excellent diagrams and illustrations are included that illustrate concepts of motor control by hierarchical reflex integration. Experiments that illustrate "functional automatisms" are well illustrated.
Assessment: The author has achieved a lucid and well-argued presentation of his model of neurocomputation. The strength of the model rests on the fact that it originally derives from biological control systems (vonHolst) and then makes use of important analogical extensions to achieve generality and complexity not found in more "mechanical" physics based models such as those of Hopfield and Tank. The author's approach thus stands as an important biological alternative to the dominant mechanistic trend in neural network theory.

4 Stars! from Doody