The Java Technology To Simulate The BP Network Of Artificial Neural Network

This paper expatiates that we use the Java technology to simulate the BP network systematically. The BP network has a very simple construction, and has the very strong ability to classify the pattern, so it is applied extensively. Making use of the Java technology to realize the simulating of the BP network, just to make use of features of the object-oriented and transplantable of Java, and Java running environment makes the simulation system works steadily. The simulation system interface is friendlier, and can be used and understood more vividly, practically.The paper firstly analyzes the possibility of the simulation system, illuminates the problems that maybe appear during the process, and brings up the possible solution to these problems. The paper introduces the principle, the model and the characteristic of artificial neural network. The network has the ability of learning, it can only be inputted a series of examples (samples), and then the network can acquire the control relation of the data from the samples. According to the BP arithmetic, we draw out the BP flow chart, and explain features of BP network. As a used widely and most excellent programming language, Java has many features that other programming languages do not possess, such asindependency, distributing and security, so it makes the application programs distributing simpler.This paper illuminates the simulation procedure of BP network by Java technique in detail, explains the basic structure of the BP network of the simulation system and the description ability of Java which as the facility of describing the artificial neural network. Especially it describes the object-oriented structure, and the object-oriented technique of Java, introduces the using flow chart. By testing, we draw out the impression picture. The simulation system possesses features as follow: used easily, strong function, validity, extensibility, high-powered and security.