Based on survey data, the migrant workers of Chengdu-Chongqing area are divided into fourclasses by K-means Clustering method. The first type：the challenge-like ones；The second type：high burden,dovish ones；The third type：potential and radical ones (the extreme of the first type)；The forth type：em-ployment disadvantage, a heavy burden (the extreme of the second type).We find that age, education level,stress of capital sources, motivation of employment have a great influence on job selecting, expected salaryand employment area selection.
展开▼
