Studies on incidence pattern recognition based on information entropy

Abstract

For pattern recognition, the weight of the feature index is very important and is usually used to measure the feature index's importance. The weight is usually divided into two types. One is determined by the knowledge and experience of experts or individuals, and called subjective weight; the other is based on statistical properties and measurement data, and is called objective weight. In this paper, a new objective weight, information entropy weight (IEW) is defined and constructed based on information entropy and the practical background of the survey data. On the basis of gray relation analysis and IEW presented here, a new concept of information incidence degree (IID) is proposed. A new method of incidence pattern recognition based on IID is set up, and applied to soil nutrient data processing. The results of simulation application show that the method presented here is feasible and effective. It provides a new research approach for information pattern recognition.