论文部分内容阅读
通过支持向量机(SVM)对客车车型的长,宽,高,宽长比等7个特征进行特征选择,得到的准确率最高的子集是长、宽、高、宽长比、宽高比,以它作为样本特征进行分类.对客车的4类车型进行分类,每类车型选择80个样本,50个样本进行训练,30个样本进行预测,结果表明:对1类车型的分类准确率可达到100%,对2类和4类车型可达到96%以上,对3类车可达到93%以上.得到了比选用长、宽、高作为特征进行分类更优的结果.然后运用加入参数寻优的SVM对客车的4类车型进行分类,并加以比较.基于高斯函数的特性,两次用到SVM进行机器学习时,核函数均选用RBF核函数.
The feature selection of seven features such as the length, width, height, width-length ratio and so on of the bus models by the support vector machine (SVM), the subset with the highest accuracy is the length, width, height, aspect ratio, aspect ratio , And classify them as sample features.According to the four types of buses, 80 samples for each type of vehicle, 50 samples for training, and 30 samples for prediction are used to predict the classification accuracy of Class 1 models Up to 100%, up to 96% for category 2 and 4 models and over 93% for category 3 vehicles, resulting in a better classification than using the features of length, width and height, and then adding the parameters The excellent SVM classifies the four types of buses and compares them.According to the characteristics of Gaussian function, when RBM is used for machine learning twice, RBF kernel function is selected as kernel function.