Progressive trends on the application of artificial neural networks in animal sciences - A review

Abstract

In recent years, artificial neural networks have become the subject of intensive research in a number of scientific areas. The high performance and operational speed of neural models open up a wide spectrum of applications in various areas of life sciences. Objectives pursued by many scientists, who use neural modelling in their research, focus - among others - on intensifying real-time calculations. This study shows the possibility of using Multilayer-Perceptron (MLP) and Radial Basis Function (RBF) models of artificial neural networks for the future development of new methods for animal science. The process should be explained explicitly to make the MLP and RBF models more readily accepted by more researchers. This study describes and recommends certain models as well as uses forecasting methods, which are represented by the chosen neural network topologies, in particular MLP and RBF models for more successful operations in the field of animals sciences.

Keywords: livestock animal science; machine learning; machine models.

Figure 1. Scheme of an artificial neuron as a basic element of the artificial neural network (Korbicz et al. 1994)

Figure 2. The structure of an example of an artificial neural network – Multilayer Perceptron (Tadeusiewicz 1993)

Figure 3. Structure of a single layer MLP neural network – example (Tadeusiewicz 1993)

Figure 4. Structure of an RBF neural network – example (Tadeusiewicz 1993)