Hazard / survival modeling is often under-applied given its broad use cases. For example, churn prediction is often posed as a classification problem (did churn or not), when the time component is often given short shrift (when, if ever did the churn happen?)
We hope to argue that hazard modeling is a better fit for these types of problems; spread general awareness of survival modeling, metrics, and data censoring; and describe how Opendoor uses these models to estimate our holding times for homes and mitigate risk, detailing scalability and other technical challenges we had to overcome.
David is a data scientist on the pricing and revenue optimization group at Opendoor in San Francisco. The team focuses on developing per-home liquidity estimation and building responsive pricing models.Prior to joining Opendoor, David was a data scientist at Rdio working on music recommenders. He holds a BS in computer science from Binghamton University, and an MS in computer science from the University of Illinois.
Xinlu is a data scientists on the pricing and revenue optimization group at Opendoor in San Francisco. The team focuses on developing per-home liquidity estimation and building responsive pricing models.Before joining Opendoor, Xinlu was a theoretical particle physicist at Stanford. She holds a Master of Music in piano performance from Peabody Conservatory and PhD in physics from Stanford.