|
Book Review |
M. Alvarez Grima. Balkema, Rotterdam, 2000. EUR 83.00 hardback; viii+244pp. ISBN 90-5809-337-9.
| The first 250 words of the full text of this article appear below. Images appear only in PDF or full-text views. |
As an introduction to the world of fuzzy modelling and neural networks, this volume is to be commended. The title refers to engineering geology but the reader will find that the applications are in rock mechanics, concerned largely with rock strength in relation to machine cutting and boring. A short case history at the end, entitled geological mapping, is actually concerned with geophysical data processing and their integration utilizing a neural net. Conventional geological mapping principles and documentation are not considered.
The book is structured as follows, with an Introduction (Ch. 1) providing a readable and succinct account of the soft computing approach, helpfully drawing attention to the strengths and weaknesses of the modelling tools that are considered. Subsequent chapters describe these tools in more detail (Ch. 2 and Ch. 3) and then consider a number of applications: to rock cutting (Ch. 4), tunnel boring machines (Ch. 5), rock strength (Ch. 6) and mapping (Ch. 7).
The quantiative approach which the neuro-fuzzy methodology promises for the consideration of qualitative descriptions and judgmental assertions cannot avoid the uncertainties in the original observations, for instance recorded as field data. Yet uncertainty is not addressed here, so caution needs to be applied to avoid believing that word-based descriptions can somehow be ‘improved upon’ by the application of this new methodology.
The geological mapping example appears to be a planar (2-D) application, and as such would have limited applicability in crust affected by faulting or significant tectonic disturbance. The reader is left thinking: So
