摘要
支持向量机增量学习方法是在回归支持向量机的基础上,在加入新增样本时有效利用历史训练结果,避免样本的重复训练,得到较准确的分类结果,其回归预测优于传统方法。工业污染预测能够及时预测工业污染物的变化,有效防止污染事故的发生。将一种改进的支持向量机增量学习方法用于工业污染预测中,通过实验结果证明:支持向量机增量学习能够较准确地预测废气污染浓度变化趋势,为工业污染预测提供了新的方法。
Based on regression support vector machine （SVM）, SVM incremental learning method effectively uses the results of historical training to obtain more accurate classification results, avoiding repetitive training samples, and its regression forecast is better than that of the traditional method. Industrial pollution forecast is able to predict changes in the industrial pollutants in a timely manner, effectively preventing the occurrence of pollution accidents. An improved SVM incremental learning method is used in industrial pollution forecasting. The experimental results and their analysis show preliminarily that SVM incremental learning method can more accurately predict the concentration of exhaust pollution trends. It provides a new method of industrial pollution warning.
出处
《西北工业大学学报》
EI
CAS
CSCD
北大核心
2014年第1期44-48,共5页
Journal of Northwestern Polytechnical University
关键词
支持向量机
工业污染预测
增量学习
classification, efficiency, Lagrange multipliers, MATLAB, machines
incremental learning, industrial pollution prediction optimization, pollution, support vector
