TechPrufer
Get started
Netflix
Netflix logoNetflix · Feature trail

Out-of-Memory Kill Prediction

1 post · Jul 2022

What this trail covers

Netflix is evolving its mobile application's reliability by formulating out-of-memory (OOM) kill prediction as a machine learning problem. This involves developing models to proactively identify and mitigate conditions that lead to OOM kills, thereby improving app stability and user experience.

All Netflix trails
  1. MediumThis post details the formulation of OOM kill prediction as a machine learning problem. It describes the data collection process, feature engineering, model selection (e.g., gradient boosted trees), and evaluation metrics used to predict OOM kills on the Netflix app. The goal is to reduce OOM kills by understanding their root causes and developing predictive capabilities.