Investigating Risk Factors Affecting the United States Teenage Pregnancy Rates based on Support Vector Machine Classifier

Teenage pregnancy remains an important issue in the United States. After several years of decline, the teenage pregnancy rate is rising with a birth rate of 41.9 per 1,000,000 live births for mothers between the ages of 15 and 19 (Martin, 2007). The purpose of this study is to investigate important risk factors affecting the teenage pregnancy rate. Through a review of the literature we were able to identify important factors including state per capita income, state unemployment rate, state poverty rate, percent of state population with less than nine years of education, and percent of state population with a high school diploma. The data used for this study comes from the U.S. Department of Labor and was representative of fifty states between the years 2000 and 2006. Support vector machine was used to establish the nonlinear relationship between teenage pregnancy rate and the risk factors. The SVM classifier maps input data from the input space into the high-dimensional feature space, and seeks an optimal hyperplane to separate data from opposite multiple classes. The results show that the polynomial SVM model yields the most accurate results with an R-squared value of 0.70. The top three ranking contributors to the teenage pregnancy rate are state per capita income, state poverty rate, and percent of state population who entered high school but did not receive a diploma. This finding highlights the need for prevention programs among these high risk populations and the role socioeconomic status and education play in teenage pregnancy.