Abstract: The production of the byproduct in natural gas enterprise is greatly affected by the market vibration，the forecast of the byproduct production usually relies on the experience of the people concerned.The amount of work is large，but the efficiency is relatively low.The statistical analysis method and the traditional machine learning method have some limitations in the application in this field.So a new machine learning method—Support Vector Machine，in the background of the enterprise is introduced.It can solve the forecast problem by means of Support Vector Machine method.It is proved that by using the support vector machine，the enterprise’s meet can be met，and it can also produce satisfactory result.

Key words: Support Vector Machine（SVM）, statistical learning theory, Principal Components Analysis（PCA）, forecast of production

摘要： 天然气企业副产品产量受市场的波动影响很大，对于副产品产量的预测，通常都是依靠工作人员的经验，工作量大，效率不高。回归统计分析方法及传统的机器学习方法在这个领域的应用存在着一些局限性，介绍一种新的机器学习算法—支持向量机，以企业为背景，运用支持向量机算法来解决预测问题，实验证明采用支持向量机能够满足企业的需求，得到满意的效果。

关键词: 支持向量机, 统计学习理论, 主成分分析, 产量预测