Using support vector machine for online purchase predication
Xiaoman Liu, Jing Li · 2016
E-commerce has become a vital contributor to China's national economy. A mass of users' behavioral data on E-commerce platforms such as browse, click and purchase have being accumulated during DT era. Using machine learning algorithms to explore patterns behind big data grows into a new focus of research. In this paper, firstly, we use SQL Server to do feature extraction on those behavioral data. Secondly, Libsvm, a software package based on SVM, is used to train the features collected above to build a predicting model. Finally, we employ the model to predict the future buying conditions of online consumers and acquire a desirable outcome. Thus, to a certain extent, our study has a practical significance to discover regular patterns of online shopping as well as improve products recommendation accuracy and conversion rate of e-commerce.