Abstract

   This paper presents a Java implementation of a feed forward neural network, used for the ranking of a working place. We took into account the following input parameters: temperature, humidity, noise and luminosity. These parameters determine the degree of comfort of a working place. Because the relationship between those four parameters and the output result (workplace ranking) is non-linear, a neural network is a very good tool to study this ranking. A neural network has the ability to quantify even non-evident relationships between input parameters as far as we have a set of good training examples.