The Use of Artificial Neural Networks (ANNs) in Aquatic Ecology

Encyclopedia of …, 2009

Zhang WJ. Computational Ecology: Artificial Neural Networks and Their Applications. World Scientific, Singapore, 2010

Due to the complexity and non-linearity of most ecological problems, artificial neural networks (ANNs) have attracted attention from ecologists and environmental scientists in recent years. As these networks are increasingly being used in ecology for modeling, simulation, function approximation, prediction, classification and data mining, this unique and self-contained book will be the first comprehensive treatment of this subject, by providing readers with overall and in-depth knowledge on algorithms, programs, and applications of ANNs in ecology. Moreover, a new area of ecology, i.e., computational ecology, is proposed and its scopes and objectives are defined and discussed. Computational Ecology consists of two parts: the first describes the methods and algorithms of ANNs, interpretability and mathematical generalization of neural networks, Matlab neural network toolkit, etc., while the second provides case studies of applications of ANNs in ecology, Matlab codes, and comparisons of ANNs with conventional methods. This publication will be a valuable reference for research scientists, university teachers, graduate students and high-level undergraduates in the areas of ecology, environmental sciences, and computational science. Contents: Artificial Neural Networks: Principles, Theories and Algorithms: Feedforward Neural Networks Linear Neural Networks Radial Basis Function Neural Networks BP Neural Network Self-Organizing Neural Networks Feedback Neural Networks Design and Customization of Artificial Neural Networks Learning Theory, Architecture Choice and Interpretability of Neural Networks Mathematical Foundations of Artificial Neural Networks Matlab Neural Network Toolkit Applications of Artificial Neural Networks in Ecology: Dynamic Modeling of Survivor Process Simulation of Plant Growth Process Simulation of Food Intake Dynamics Species Richness Estimation and Sampling Data Documentation Modeling Arthropod Abundance from Plant Composition of Grassland Community Pattern Recognition and Classification of Ecosystems and Functional Groups Modeling Spatial Distribution of Arthropods Risk Assessment of Species Invasion and Establishment Prediction of Surface Ozone Modeling Dispersion and Distribution of Oxide and Nitrate Pollutants Modeling Terrestrial Biomass Readership: Research scientists, university teachers, graduate students and high-level undergraduates in the area of ecology, environmental sciences and computational science.

Bio-Algorithms and Med-Systems, 2015

Neural networks become very popular as a too] for modeling of numerous systems, including technological , economical, sociological, psychological, and even political ones. On the contrary neural networks are models of neural structures and neural processes observed in a real brain. However, for modeling of real neural structures and real neural processes occurring in a living brain, neural networks are too simplified and too primitive. Nevertheless , neural networks can be used for modeling the behavior of many biological systems and structures. Such models are not useful for explanation, taking into account the biological systems and processes, but can be very useful for the analysis of such system behavior, including the prognosis of future results of selected activities (e.g. the prognosis of results of different therapies for modeled illnesses). In this paper, selected examples of such models and their applications are presented.

Artificial intelligence has been the inspiration and goal of computing since the discipline was first conceived by Alan Turing. Our understanding of the brain has increased in parallel with the development of computers capable of modelling its functions. While the human brain is vastly complex, too much so for the computation abilities of modern super computers, interesting results have been found while modelling the nervous system of smaller creatures such as the salamander [3].

ecological modelling, 2007

Neuroet is an easy-to-use artificial neural network (NN) package designed to assist with determining relationships among variables in complex ecological and biological systems. The package, which is available for download from the web site http://noble.ce.

Artificial Neural Network (ANN) has emerged with advancement of Information and Communication technology and biological sciences during last decades. The aim is to utilize technology and construct machines that will work like brain of humans. The internal architectural requirements of such a machine is to have huge simultaneous memory and storage in consistent with intensive processing power to cater the ambiguous information and behave like human brain. ANN has broad range of applications in today's business and IT industry. This paper aims to investigate the working of ANN and its applications in real environment.

2009

Phytoplankton biomass within the Saginaw Bay ecosystem (Lake Huron, Michigan, USA) was characterized as a function of select physical/chemical indicators. The complexity and variability of ecological systems typically make it difficult to model the influences of anthropogenic stressors and/or natural disturbances. Here, Artificial Neural Networks (ANNs) were developed to model chlorophyll a concentrations, a measure for water-column phytoplankton biomass and a proxy for system-level health. ANNs act like "black boxes" in the sense that relationships are encoded as weight vectors within the trained network and as such, cannot easily support the generation of scientific hypotheses unless these relationships can be explained in a comprehensible form. Accordingly, the 'knowledge' and/or rule-based information embedded within ANNs needs to be extracted and expressed as a set of comprehensible 'rules'. Such extracted information would enhance the delineation and understanding of ecological complexity and aid in developing usable prediction tools. Comparisons of various computational approaches (including TREPAN, an algorithm for constructing decision trees from neural networks) used in extracting rule-based information from trained Saginaw Bay ANNs are discussed.

In this research project, the features of biological and artificial neural networks were studied by reviewing the existing works of authorities in print and electronics on biological and artificial neural networks. The features were then assessed and evaluated and comparative analysis of the two networks was carried out. The metrics such as structures, layers, size and functional capabilities of neurons, learning capabilities, style of computation, processing elements, processing speed, connections, strength, information storage, information transmission, communication media selection, signal transduction and fault tolerance were used as basis for comparison. A major finding in the research showed that artificial neural networks served as the platform for neuro-computing technology and as such a major driver of the development of neuron-like computing system. It was also discovered that Information processing of the future computer systems will greatly be influenced by the adoption of artificial neural network model.

The first few pages of any good introductory book on neurocomputing contain a cursory description of neurophysiology and how it has been abstracted to form the basis of artificial neural networks as we know them today. In particular, artificial neurons simplify considerably the behavior of their biological counterparts. It is our view that in order to gain a better understanding of how biological systems learn and remember it is necessary to have accurate models on which to base computerized experimentation. In this paper we describe an artificial neuron that is more realistic than most other models used currently. The model is based on conventional artificial neural networks (and is easily computerized) and is currently being used in our investigations into learning and memory.

umerous advances have been made in developing intelligent systems, some inspired by biological neural networks. N Researchers from many scientific disciplines are designing artificial neural networks (A"s) to solve a variety of problems in pattern recognition, prediction, optimization, associative memory, and control (see the "Challenging problems" sidebar).

Lecture Notes in Computer Science, 2005

We characterize the first hardware implementation of a self-organizing map algorithm based on axon migration. A population of silicon growth cones automatically wires a topographic mapping by migrating toward sources of a diffusible guidance signal that is released by postsynaptic activity. We varied the diffusion radius of this signal, trading strength for range. Best performance is achieved by balancing signal strength against signal range. Comments Comments

Geosciences

What is the optimal level of chaos in a computational system? If a system is too chaotic, it cannot reliably store information. If it is too ordered, it cannot transmit information. A variety of computational systems exhibit dynamics at the “edge of chaos”, the transition between the ordered and chaotic regimes. In this work, we examine the evolved neural networks of Polyworld, an artificial life model consisting of a simulated ecology populated with biologically inspired agents. As these agents adapt to their environment, their initially simple neural networks become increasingly capable of exhibiting rich dynamics. Dynamical systems analysis reveals that natural selection drives these networks toward the edge of chaos until the agent population is able to sustain itself. After this point, the evolutionary trend stabilizes, with neural dynamics remaining on average significantly far from the transition to chaos.

arXiv (Cornell University), 2023

As our understanding of the mechanisms of brain function is enhanced, the value of insights gained from neuroscience to the development of AI algorithms deserves further consideration. Here, we draw parallels with an existing tree-based ANN architecture and a recent neuroscience study [27] arguing that the error-based organization of neurons in the cerebellum that share a preference for a personalized view of the entire error space, may account for several desirable features of behavior and learning. We then analyze the learning behavior and characteristics of the model under varying scenarios to gauge the potential benefits of a similar mechanism in ANN. Our empirical results suggest that having separate populations of neurons with personalized error views can enable efficient learning under class imbalance and limited data, and reduce the susceptibility to unintended shortcut strategies, leading to improved generalization. This work highlights the potential of translating the learning machinery of the brain into the design of a new generation of ANNs and provides further credence to the argument that biologically inspired AI may hold the key to overcoming the shortcomings of ANNs. * Contributed equally. Preprint. Under review.

An artificial neural network (ANN)-based technology e a 'Grey-Box', originating the iterative selection, depiction, and quantitation of environmental relationships for modeling microalgal abundance, as chlorophyll (CHL) a, was developed and evaluated. Due to their robust capability for reproducing the complexities underlying chaotic, non-linear systems, ANNs have become popular for the modeling of ecosystem structure and function. However, ANNs exhibit a holistic deficiency in declarative knowledge structure (i.e. a 'black-box'). The architecture of the Grey-Box provided the benefit of the ANN modeling structure, while deconvolving the interaction of prediction potentials among environmental variables upon CHL a. The influences of (pairs of) predictors upon the variance and magnitude of CHL a were depicted via pedagogical knowledge extraction (multi-dimensional response surfaces). This afforded derivation of mathematical equations for iterative predictive outcomes of CHL a and together with an algorithmic expression across iterations, corrected for the lack of declarative knowledge within conventional ANNs. Importantly, the Grey-Box 'bridged the gap' between 'white-box' parametric models and black-box ANNs in terms of performance and mathematical transparency. Grey-Box formulations are relevant to ecological niche modeling, identification of biotic response(s) to stress/disturbance thresholds, and qualitative/quantitative derivation of biota-environmental relationships for incorporation within stand-alone mechanistic models projecting ecological structure.