Graph networks (Relational Neural Networks)

The method “generalizes and extends various approaches for neural networks that operate on graphs, and provides a straightforward interface for manipulating structured knowledge and producing structured behaviors”. Tipul resursei: model/metodă; Scop: raționament neural care lucrează cu relații (în grafuri) în loc de caracteristici.

Producător: DeepMind

  1. Battaglia PeterW., Pascanu Razvan , Lai Matthew, Rezende Danilo, Kavukcuoglu Koray(2016): Interaction Networks for Learning about Objects, Relations and Physics.
  2. Battaglia et al. (2018), Relational inductive biases, deep learning, and graph networks =>https://arxiv.org/pdf/1806.01261.pdf
  3. Kipf Thomas (2016), Graph convolutional networks: http://tkipf.github.io/graph-convolutional-networks/

Leave a Reply

Your email address will not be published. Required fields are marked*