a new area of Machine Learning research concerned with the technologies used for learning hierarchical representations of data, mainly done with deep neural networks (i.e. networks with two or more hidden layers), but also with some sort of Probabilistic Graphical Models.
What is Deep Learning?
Deep Learning is an area of machine-learning which attempts to build neural-networks to learn complex functions by using special architectures composed of many layers (hence the term "deep").
Deep architectures allow more complex tasks to be learned because, in addition to these neural networks having more layers to perform transformations, the larger number of layers and more complex architectures of the neural network allow a hierarchical organization of functionality to emerge.
Deep Learning was introduced into machine learning research with the intention of moving machine learning closer to artificial intelligence. A significant impact of deep learning lies in feature learning, mitigating much of the effort going into manual feature engineering in non-deep learning neural networks.
New to Deep Learning?
There are a variety of resources including books, tutorials/workshops, etc. for those looking to learn more about Deep Learning.
A popular introductory tutorial is:
SciPy 2020 Conference Tutorial:
Some popular introductory books:
Deep Learning with Python, by François Chollet
Deep Learning with PyTorch, by Eli Stevens, Luca Antiga, and Thomas Viehman
Deep Learning, by Ian Goodfellow
Resources
Papers
Books
Neural Networks and Deep Learning By Michael Nielsen - this is a free book with associated Python source code on Github
Deep Learning Made Easy with R: A Gentle Introduction For Data Science
Videos
- Neural Networks Demystified - accompanied by a set of Jupyter Notebooks
- Deep Learning by Andrew Ng
Stack Exchange Sites
Other StackExchange sites with Deep Learning tag: