Should ensemble members be calibrated? | AISC

Channel:
Canada LLMs Explained - Aggregate Intellect - AI.SCIENCE
Subscribers:
22,600
Published on ● Video Link: https://www.youtube.com/watch?v=tXNvM8W3eA8


Duration: 50:24
136 views
6

For slides and more information on the paper, visit https://ai.science/e/should-ensemble-members-be-calibrated--zafar-2021-01-12

Speaker: Xixin Wu; Discussion Facilitator: Nour Fahmy

Motivation:
Underlying the use of statistical approaches for a wide range of applications is the
assumption that the probabilities obtained from a statistical model are representative
of the “true” probability that event, or outcome, will occur. Unfortunately, for
modern deep neural networks this is not the case, they are often observed to
be poorly calibrated.



Other Videos By LLMs Explained - Aggregate Intellect - AI.SCIENCE


2021-03-10Modeling Dissolution of Compact Planetary Systems
2021-03-09High Dimensional Inference in the Universe
2021-03-08BERT & NLP Explained
2021-03-08Introducing AI BRIEFS
2021-03-04Deep Unsupervised Learning for Climate Informatics
2021-03-03New Directions in Computer Vision
2021-03-02Machine Learning Product Management in the Insurance Industry
2021-03-01Running ML Experiments on the Edge
2021-02-26Data exploration – can machine learning discover new drugs? | AISC
2021-02-25Building a Quantum Computer with Trapped Ions
2021-02-25Should ensemble members be calibrated? | AISC
2021-02-25Topological Deep Learning
2021-02-24Explainable AI to Automate Loan Underwriting Decisions | AISC
2021-02-23'Less Than One'-Shot Learning (author speaking)
2021-02-18Machine Learning in Mobile Cybersecurity: An Overview
2021-02-17Author speaking: Proper Machine Learning Explanations through LIME using OptiLIME framework | AISC
2021-02-11Non-Euclidean Universal Approximation | AISC
2021-02-09Explainable AI with Layer-wise Relevance Propagation (LRP)
2021-02-04An Introduction to Quantum Computing
2021-02-04Predicting compound activity from phenotypic profiles
2021-02-03A Survey on the Explainability of Supervised Machine Learning