Survival Random Forest to Predict Time to Fill

Abstract

Traditionally, the time-to-fill metric is used as a scorecard for past performance. An organization may use time to fill to assess the performance of its internal recruiting team, or as a way to set service level agreements with outsourced recruiting partners. By first developing a set of quantifiable job features and then applying survival analysis to historical time-to-fill data, we build a predictor to assess the probability a job will remain open beyond its target time-to-fill date, enabling us to commit additional resources to high risk jobs at the beginning of the recruiting process.