Advancing High School Dropout Predictions Using Machine Learning

The importance of high school completion for jobs and postsecondary opportunities is well-documented. Schools, districts, and states are increasingly concerned about improving outcomes for vulnerable student populations. Combined with U.S. federal laws where high school graduation rate is a core performance indicator, states face pressure to actively monitor and assess high school completion. This study employs machine learning algorithms to preemptively identify students at-risk of dropping out of high school. We leverage North Carolina statewide administrative data to build an early warning prediction model that identifies students at-risk of dropping out and groups them based on similar patterns and characteristics. This study provides guidance and informed knowledge about how districts and states can capitalize "big data" to identify vulnerable students preemptively.