| Summary: | Support Vector Machines (SVMs) are excellent tools in the field of machine learning that are well suited for classification and regression problems. However, SVM is highly sensitive to data. In recent years, many researchers have continued to use various data preprocessing methods to improve model performance. Nevertheless, there is a lack of comparison of multiple datasets preprocessing methods, and there is also a lack of experiments on classification targets. This research aims to compare the performance of different preprocessing methods by using vector machines with radial basis function (RBF) kernel support. Three data sets of pedestrians, vehicles, and traffic signs are classified, and four common data preprocessing methods are used: data cleaning, feature selection, feature scaling, and dimensionality reduction. Observing the results through the use of assessment metrics accuracy, precision, recall, and F1 score. The experiment shows that the preprocessing method of feature selection is superior to the other three preprocessing methods. It achieves an accuracy of 0.985, a precision of 0.986, a recall rate of 0.985 and an F1 score of 0.986. This research has important research significance for automatic driving target classification. © 2024 IEEE.
|
|---|