Robert Lange on NN Pruning and Collective Intelligence
Machine Learning Street Talk (MLST)
00:00
Unpacking the Lottery Ticket Hypothesis
This chapter explores the lottery ticket hypothesis, revealing how overparameterization in deep networks facilitates escape from local minima. It discusses strategies for training sparse neural networks by identifying winning subnetworks and highlights significant open questions that warrant further investigation in neural network optimization.
Transcript
Play full episode
