From the course: Learning Graph Neural Networks
Join today to access over 24,400 courses taught by industry experts.
Challenges of using graphs in machine learning
From the course: Learning Graph Neural Networks
Challenges of using graphs in machine learning
- [Instructor] Before we discuss how graph neural networks work, let's talk about the challenges of using graphs as inputs to machine learning models. Before we discuss graphs, let's consider a machine learning problem that we are familiar with, image classification using convolutional neural network. We represent images as multidimensional tensors, we pass the images through the layers of the convolutional neural network to perform image classification. We usually think of images as rectangular grids and represent them as arrays. Another way to think of images is as graphs with a regular structure where each pixel represents a node and every pixel is connected via an edge to every adjacent pixel. Images tend to be fairly straightforward graphs because they have a regular grid-like structure where each pixel is consistently positioned relative to its neighbors, making it fairly simple to apply uniform operations like convolution across the entire image. In contrast, graphs tend to…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
