Deep Learning Insights
Diogo discusses the impressive modularity of deep learning and its potential to solve any computable problem, emphasizing that practical challenges like data efficiency often hinder progress. He highlights the flaws in commonly used datasets, which can lead to overfitting and reproducibility issues, stressing the importance of understanding these limitations for future advancements in the field.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Deep Learning: Modular in Theory, Inflexible in Practice with Diogo Almeida - #8
Related Questions