The Bayesian Network In The Agricultural Experts Within Application Systems Research

The agricultural expert system is applied to agricultural production, to increase agricultural production efficiency and to promote agricultural knowledge to modernize agriculture has great significance. The agricultural sector in general knowledge is an uncertainty, in order to establish·highly effective agricultural expert systems, so this article by express and Bayesian network going to handle uncertainty. This is based on the way the rules, semantic a clear and two-way reasoning, so quick to debug and reconstructions, etc.This paper mainly introduces and to create Bayesian network based on the agricultural expert system arising from the two key issues are Knowledge library build (the field of knowledge according to establish the corresponding Bayesian network) and Bayesian network reasoning algorithm is going to the realization of that evidence given in the examination of probability. In view of practical applications, based on the paper discussed the establishment of a Bayesian network agricultural systems and establish an expert how to trounce Bayesian network structure and the net every tabulate the table. The conditional probability create network to reduce the number of conditional probability, the use of a "technology" and try to separate simple structure, thereby reducing complicated problem. In the calculation that the "local structures" in less that number for conditional probabilities table. the algorithm for this reason, the same reasoning of the fundamental algorithm Gibbs sampling and accurate inferences in terms of the cutset conditioning and the Bayesian network algorithms, on the basis of the decision to use the net. The decision of the expert system is used in agriculture.At last Bayesian network policy-based agricultural experts system:Shuan chilies planted system, the policy decisions and Shuan chilies the disease diagnosed system. Through Decisions taken by the network can get kind of measures to make for Shuan chilies maximize the benefit. The results and domain experts the result basically unanimous. This explains the Bayesian network to deal with agriculture and the uncertainty of the knowledge is more efficient.The main contribution following: Expert system for agricultural applications research, and have some more examples of success. If only by this single approach to agriculture in the uncertainty of the complex, is very far from adequate. based on the handle of inadequate, the text adopted Bayesian network to express the uncertainty of knowledge and processing, this will increase the accuracy of the expert system of reasoning and the uncertainty of that scope.Bayesian network is going to handle uncertainty knowledge of the best tool. The article provides an overview of the uncertainty of knowledge and treatment, and illustrates Bayesian network grammar and semantics of Pompeii and properties, etc. The express, reasoning and learning are the three big Bayesian network study domain, the three areas has made some important progress, but there are still a lot of work to do. This article in the signify and reasoning of Bayesian network done some work.Based on the Bayesian network establish an expert system arising from the question, this is the key to Bayesian network build, use "separate technology "and "local structures" and transform existing deterministic Knowledge factor for simplified Bayesian network would build, and Bayesian network applied to agricultural experts to receive a very good results. On reasoning, except for the introduction and implementation has the same reasoning Gibbs in the sampling method, it focuses on the precise and the reasoning of the set of conditions (cutset conditioning) algorithm. And give a new algorithm based on reasoning of the decision of the methods and procedures. On this basis, according to the development of policy-based Bayesian network of production forecast to the Shuan chilies, the Shuan chilies to policy decisions and decisions of the disease diagnosis expert system.