From the course: MLOps with Databricks
Join today to access over 25,500 courses taught by industry experts.
MLOps principles - Databricks Tutorial
From the course: MLOps with Databricks
MLOps principles
- [Presenter] Companies rely on machine learning models as part of their business activity, and the real business impact is involved if model predictions are not unavailable or of bad quality. There are enough examples out there of how things went wrong. Zillow, a real estate company, lost over $500 million in 2021 due to houses bought at a too high price because of inaccurate machine learning predictions. The goal of MLOps is to minimize the risks of things going wrong. This brings us to the main four MLOps principles: documentation, code quality, traceability and reproducability, monitoring and alerting. No one likes documenting things, but also no one likes losing the time trying to figure out why things were implemented in a certain way. It's crucial to document business goals, KPIs, data definitions, ML system architecture, and the choice of machine learning models, and also other relevant information. In this way, information stays if there are changes in the team. Bad code…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.