Data Sampling Using Data Maps
summary
Shows how to quantify each training example’s influence on model learning using data maps (metrics like confidence, variability, and correctness) and then use those measures for smarter data sampling. Demonstrates applying these techniques to machine learning training pipelines (no additional technologies beyond standard ML tooling are specified).
AI-written summary · about
related talks
31mLife cycle of data becoming a model
39mDoing Data: The Critical Process of Data Preparation (...)
01h 33mUnderstanding and diagnosing your machine-learning models
30mMachine Learning Techniques for Class Imbalances & Adversaries
41mModels Schm-odels: Why you Should Care about Data-Centric AI
27mA Look Inside Applied Machine Learning
37mMachine Learning, The Right Way
34mPanel 02 - Data Quality & Feature Stores