A repository of beginner resources to some of the recently developing fields in deep learning that we are exploring, to get a basic understanding of the field.
Please feel free to contribute to this repository by making a pull request!.
The field mainly focused on applying deep learning on non euclidean data like graphical data using architectures like Graph Convolutional Networks, Graph Neural Networks etc.
- A Comprehensive Survey on Graph Neural Networks [Link]
- GRAPH CONVOLUTIONAL NETWORKS - Blog post by Thomas Kipf [Link]
- PyTorch Geometric - A framework based on pytorch for handling graphical data [Link]
- DEEP GRAPH LIBRARY - Another popular framework for graphical neural networks [Link]
A highly active area of recent research, which aims to learn models which can be easily adpated to learn new tasks often said as "learning to learn".
- CS 330- Deep Multi-Task and Meta Learning By Chelsea Finn, Stanford [Link]
- Meta-Learning: Learning to Learn Fast - Blog post in Lil'Log [Link]
Deals with problems in sequential decision making, has a lot of applications in robotics and simulations.
- Reinforcement Learning Course by David Silver [Link]
- CS 285 UC BERKLEY - Deep Reinforcement Learning course by Sergey Levine [Link]
- Open AI Spinning Up in Deep RL - Good for beginner implementation details [Link]
Deals with problem of expensive data labelling and tries to figure out which unlablled data points to label so as to bring maximum improvement to model performance.
- Overview of Active Learning for Deep Learning- Highly intuitive blog post for basics of Active Learning [Link]