Disaggregated model evaluation and comparison

Channel:
Microsoft Research
Subscribers:
343,000
Published on ● Video Link: https://www.youtube.com/watch?v=NYXRrLzGiFk


Duration: 18:17
451 views
22

This video is an introduction to disaggregated model evaluation and comparison in Machine Learning. Throughout this video we will describe the main concepts of disaggregated evaluation, how these concepts relate to Responsible AI approaches of scoring and evaluating a ML model, and how they can be used in practice.

Adaptive Systems and Interaction Group: https://www.microsoft.com/en-us/research/group/adaptive-systems-and-interaction/



Other Videos By Microsoft Research


2023-04-27Innovating through uncertainty: Getting super curious and combining disparate elements
2023-04-13WiDS Career Panel: Gabriela de Queiroz, Juliet Hougland (Netflix), and Samantha Sifleet
2023-03-24Learning to Exploit Temporal Structure for Biomedical Vision-Language Processing
2023-03-23Foundation models and the next era of AI
2023-02-24Behind the label: Glimpses of data labelling labours for AI
2023-02-17Art of doing disruptive research
2023-02-17Fighting the Global Social Media Infodemic: from Fake News to Harmful Content
2023-02-15Responsible AI Tracker Tour
2023-02-14Automating Commonsense Reasoning
2023-02-13Reinforcement Learning (RL) Open Source Fest 2022 Final Project Presentations
2023-02-13Disaggregated model evaluation and comparison
2023-02-13Neural Interfaces - Towards a new generation of human-computer interface
2023-02-13Galea: The Bridge Between Mixed Reality and Neurotechnology
2023-02-10Current and Future Application of BCIs
2023-02-01Seeing AI app - Creating a Route
2023-02-01Seeing AI app - Indoor Navigation
2023-02-01Seeing AI app - World Channel
2023-01-24SmartKC: A Low-cost, Smartphone-based Corneal Topographer
2023-01-11MSR-IISc AI Seminar Series: On Learning-Aware Mechanism Design - Michael I. Jordan
2022-12-22Tongue-Gesture Recognition in Head-Mounted Displays
2022-12-15Global Renewables Watch - AI for Good Lab - Geospatial