Compressed Representations in Deep Learning: From Algorithmic Information to Autoencoders and Neural Networks

Compressed Representations in Deep Learning: From Algorithmic Information to Autoencoders and Neural Networks

Learn about the fundamental concepts of compressed representations in deep learning through this 27-minute lecture from the Instituto de Física Interdisciplinar y Sistemas Complejos (IFISC), exploring the connections between algorithmic information theory and modern autoencoder architectures while discovering how compression principles influence deep learning model design and performance.