基于对应分析的训练样本的选择

本文提出一种基于对应分析的训练样本的选择方法。它从训练样本中自动地选择有代表性的典型训练样本，使得在自动分类中充分利用所采集的样本信息，以便得到满意的分类结果。通过实验与分析证明，该方法是可行的，它明显优于人工随机选择训练样本的方式。与基于Q型因子分析的训练样本选择方法相比，可以更快地得到较少的典型样本，满意的分类精度。

Training Samples Selection Method Based on Correspondence Analysis

In fact,training samples selection is very important in the classification domain and to some extent it has an influence on the classification accuracy.Q-factor analysis only considers relations among samples,but correspondence analysis makes use of feature variables and samples at the same time.Therefore,a new method about training samples selection based on correspondence analysis is proposed,which can automatically select representative and integral training samples for capturing sufficient samples information and improving classification accuracy.The experimental results demonstrate its feasibility and it can cut down computational costs and perform well compared with the manual selection of training samples and the method based on Q-factor analysis.