The Application And Research Of Architectural Ceramic Defection Based On Data Mining

Since the reform and opening, China’s architectural ceramic industry had rapid developed with global economy recovery. The ceramic production increasing and ceramic raw materials overexploitating, it will lead to the series of issues, such as low profits, low-quality products, as well as it will affect the development of the architectural ceramics. The price of Chinese architectural ceramic products is not high in international market,and the international market competition ability of chineses architectural ceramics is not strong. So the Chinese architectural ceramic interprises must improve the quality of production, reduce the cost of production, adjusting the structure of production and pay more attention to innovation.To improve the quality of architectural ceramic product and reduce the defect of architectural ceramice product, the data mining was used in the classification of architectural ceramics raw material and defects of architectural ceramics product in this paper. Discussed classification of architectural ceramics raw material, analyzed the defects of architectural ceramic products. The method provided theoretic basis and practical guidance for improving the utilization ratio of architectural ceramic raw materials, reducing architectural ceramic defects.According to the factors of architectural ceramics and the advantages of data mining, the rough sets, decision tree and the BP neural network were used in the architectural ceramics. Firstly, we introduced the three theories, and illustrated with examples. Secondly, we analysed the some defects in the theories and improved it. Finally, the architectural ceramic raw materials were classified by three improved methods, the defects of architectural ceramic were analysed. We got some experiment results.We put forward the weighted average attribute reduction algorithm and the improvement of ID3 algorithm, the model of rough set and ID3 were applied to the classification of architectural ceramic raw materials. The BP neural network was infected by the order of acceptable influence. we can make a improvement in the BP network.The model of Rough set decision tree was used in the classification of materials, and the result was very good. Rough set neural network model not only eliminated redundant noise data, also simplified the neural network’s, Rough set neural network model was applied to ceramic defect analysis, and the result is feasible, reasonable, and achieved good defect analysis results.