2021 Open Data Science
Conference

November 16 , 9:30 AM -11:00 AM PST

Machine learning models have become increasingly complex and it is imperative to utilize better tools to monitor, troubleshoot, and explain their decisions as models move from research to production environments. In this workshop, Aparna Dhinakaran, Co-Founder, CPO of Arize Al (EX-Uber Machine Learning), will discuss the state of ML production monitoring, Its challenges, and how to actively improve models in production

Learn how to validate degradation in model performance, take a deep dive to investigate the root causes of those inaccurate predictions, and set up proactive monitors to mitigate the Impact of future degradations. Experience ML observability first hand with a walkthrough of the Arize platform using practical use case examples to identify segments where your model is underperforming, troubleshoot root cause analysis, proactively monitor for future degradations, and estimate the business Impact of a models decisions.

November 16 , 1:50 PM -2:20 PM PST

An overview of Arize AI’s ML Observability platform, which helps ML teams automatically surface issues, understand and resolve why they occurred, and improve model performance continuously. Gabriel Barcelos, Senior Software Engineer at Arize AI, will demonstrate how to gain a centralized view of all your models in production and automatically monitor key performance attributes, data quality, and drift. Gabriel will leverage the Arize performance dashboard to take a step deeper into root cause analysis, and highlight feature drift and prediction drift impact. Utilize the Arize platform to surface issues easily, troubleshoot the root cause, and quickly resolve and improve your models as they go from research to production.

November 17 , 2:00 -2:45 PM PST

As more and more machine learning models are deployed into production, we must have better observability tools to monitor, troubleshoot, and explain their decisions. In this talk, Aparna Dhinakaran, Co-Founder, CPO of Arize Al will discuss the state of the commonly seen ML Production monitoring and its challenges. She will focus on using statistical distance checks to monitor features and model output in production, analyze the effects of the changes on models and use explainability techniques to determine if issues are model or data related.