Prediciton Of Product Qualified Rate Based On Support Vector Machine

With the rapid social and economic development, people’s living standards have been continuously improving. The consumers will no longer accept features and functionality of products as a passive acceptance as before, but gradually put forward their diversified and personalized requirements. At the same time, with the continuous development of globalization, driven by technological innovation and product market segments, the replacement cycle is getting shorter and shorter, competition among enterprises is is intensifying. Therefore, how to be able to do fast response to market demands, become a key factor for companies, especially manufacturing enterprises, to keep an invincible position in fierce market competition. Therefore, the many varieties of small batch production gradually replace the single piece of mass production, become the main production mode for modern enterprises.In the manufacturing process, product quality is an important indicator to measure whether the product is successful. When the mode of production changes, in order to produce products meet consumer demand, in line with relevant regulations and standards, quality management and control in the manufacturing process, especially the prediction of the rate of qualified products will also change.Combined with the characteristics of many varieties of small batch production mode, this paper adopts a support vector machine prediction model based on statistical learning theory, to predict the rate of qualified products, the contents are as follows:1) Summarize research on the prediction of the rate of qualified products, and analyze the advantages and disadvantages of various methods to provide a reference for relevant research;2) On the basis of a brief introduction and review about statistical learning theory and support vector machine theory, analyze the theoretical feasibility of using support vector machine to predict the rate of qualified products, laying a theoretical foundation for study; 3) Take account of the influence of different kernel functions, on the basis of analysis and simulation using several commonly used kernel functions, determine the nuclear function of the model used to predict in this article;4) Analyze the effect of model parameters, and make comparative analysis of the effect of several commonly used parameters selection methods, on the basis of determination of the parameters selection method, select the parameters values used to predict;5) In order to verify the prediction effect of support vector machine, this paper makes a comparison test with two other commonly used methods and support vector machine in forecasting the rate of qualified products. The results show that support vector machine prediction method can achieve relatively good prediction.