Machine Learning Methods for Ecological Applications

It is difficult to become an ecologist withou,t acquiring some breadth~ For example, we are expected to be competent statisticians and taxonomists who appreciate the importance of spatial and temporal processes, whilst recognising the potential offered by techniques such as RAPD. It is, therefore, with some trepidation that we offer a collection of potentially useful methods that will be unfamiliar, and possibly alien, to most ecologists. I don't feel old, but when I was undertaking my postgraduate research our lab calculator was mechanical. There was great excitement in my fmal year when we obtained an unbelievably expensive electronic calculator. Later I progressed to running ~obs' on a PRIME minicomputer via a collection of punched cards. Those who complain about the problems with current computers don't know how lucky they are! In 1984 I wrote a book entitled 'Computing for Biologists'. Although it was mainly concerned with writing short programs it did also look at wider aspects of the role of computers in the biological sciences. Machine learning was not mentioned in that book, probably because of ignorance but also because the methods were relatively unknown outside of the relatively small number of workers in the broad field that is now known as machine learning. During 1985 I spent a sabbatical year at York University, following their Biological Computation masters programme. This course was a unique blend of computer science, mathematics and statistics. The last 25 years have seen a tremendous growth in the application of statistical and modelling techniques to ecological problems. This expansion has been accelerated by the increasing availability of software, books and computing power. However, the suitability of some of these approaches to data analysis, in a relatively knowledge-poor discipline such as ecology, can be questioned on grounds of appropriateness and robustness. One reason for these concerns is that many ecological problems are at best poorly defined and most lack algorithmic solutions. Machine learning methods offer the potential for a different approach to these difficult problems. One definition of machine learning is that it is concerned with inducing knowledge from data, where the data could be patterns in a game of chess or patterns in the species composition of natural communities. Unfortunately ecologists have little experience of these relatively recent and novel approaches to understanding data. This is a problem that is made more complex because there is no simple taxonomy of machine learning methods and there are relatively few examples in the mainstream ecological literature to encourage exploration. This is the first text aimed at introducing machine learning methods to a readership of professional ecologists. All but one of the chapters have been written by ecologists and biologists who highlight the application of a particular method to a particular class of problem. Examples include the identification of species, optimal mate choice, predicting species distributions and modelling landscape features. A group of experienced machine learning workers, who have become interested in environmental problems, have written a chapter that demonstrates how machine learning methods can be used to discover equations that describe the dynamic behaviour of ecological systems. The final chapter reviews `real learning', offering the potential for greater dialogue between the biological and machine learning communities Front Matter....Pages i-xiii An introduction to machine learning methods....Pages 1-35 Artificial neural networks for pattern recognition....Pages 37-87 Tree-based methods....Pages 89-105 Genetic Algorithms I....Pages 107-121 Genetic Algorithms II....Pages 123-144 Cellular automata....Pages 145-183 Equation discovery with ecological applications....Pages 185-207 How should accuracy be measured?....Pages 209-223 Real learning....Pages 225-246 Back Matter....Pages 247-261