The recent success of neural networks has boosted research on pattern recognition and data mining.
Machine learning tasks, like object detection, machine translation, and speech recognition, have been given new life with end-to-end deep learning paradigms like CNN, RNN, or autoencoders.
Deep Learning is good at capturing hidden patterns of Euclidean data (images, text, videos).
But what about applications where data is generated from non-Euclidean domains, represented as graphs with complex relationships and interdependencies between objects?
That’s where Graph Neural Networks (GNN) come in, which we’ll explore in this article. We’ll start with graph theories and basic definitions, move on to GNN forms and principles, and finish with some applications of GNN.
What is a Graph?
The most fundamental part of GNN is a Graph.
In computer science, a graph is a data structure consisting of two components: nodes (vertices) and edges.
A graph G can be defined as G = (V, E), where V is the set of nodes, and E are the edges between them.
If there are directional dependencies between nodes then edges are directed. If not, edges are undirected.
Directed Graph | Source
A graph can represent things like social media networks, or molecules. Think of nodes as users, and edges as connections. A social media graph might look like this:
A graph is often represented by A, an adjacency matrix.
If a graph has n nodes, A has a dimension of (n × n).
Sometimes the nodes have a set of features (for example, a user profile). If the node has f numbers of features, then the node feature matrix X has a dimension of (n × f).
READ SOME PAPERS ABOUT GNN
👉 Top Research Papers from the ECML-PKDD 2020 Conference (on Graph Neural Networks)
Why is it hard to analyze a graph?
Graph data is so complex that it’s created a lot of challenges for existing machine learning algorithms.
The reason is that conventional Machine Learning and Deep Learning tools are specialized in simple data types. Like images with the same structure and size, which we can think of as fixed-size grid graphs. Text and speech are sequences, so we can think of them as line graphs.
But there are more complex graphs, without a fixed form, with a variable size of unordered nodes, where nodes can have different amounts of neighbors.
It also doesn’t help that existing machine learning algorithms have a core assumption that instances are independent of each other. This is false for graph data, because each node is related to others by links of various types.
Graph Neural Network
Graph Neural Networks (GNNs) are a class of deep learning methods designed to perform inference on data described by graphs.
GNNs are neural networks that can be directly applied to graphs, and provide an easy way to do node-level, edge-level, and graph-level prediction tasks.
GNNs can do what Convolutional Neural Networks (CNNs) failed to do.
Why do Convolutional Neural Networks (CNNs) fail on graphs?
CNNs can be used to make machines visualize things, and perform tasks like image classification, image recognition, or object detection. This is where CNNs are the most popular.
The core concept behind CNNs introduces hidden convolution and pooling layers to identify spatially localized features via a set of receptive fields in kernel form.
CNN on an image | Source
How does convolution operate on images that are regular grids? We slide the convolutional operator window across a two-dimensional image, and we compute some function over that sliding window. Then, we pass it through many layers.
Our goal is to generalize the notion of convolution beyond these simple two-dimensional lattices.
The insight allowing us to reach our goal is that convolution takes a little sub-patch of the image (a little rectangular part of the image), applies a function to it, and produces a new part (a new pixel).
What happens is that the center node of that center pixel aggregates information from its neighbors, as well as from itself, to produce a new value.
It’s very difficult to perform CNN on graphs because of the arbitrary size of the graph, and the complex topology, which means there is no spatial locality.
There’s also unfixed node ordering. If we first labeled the nodes A, B, C, D, E, and the second time we labeled them B, D, A, E, C, then the inputs of the matrix in the network will change. Graphs are invariant to node ordering, so we want to get the same result regardless of how we order the nodes.
Basics of Deep Learning for graphs
In graph theory, we implement the concept of Node Embedding. It means mapping nodes to a d- dimensional embedding space (low dimensional space rather than the actual dimension of the graph), so that similar nodes in the graph are embedded close to each other.
Our goal is to map nodes so that similarity in the embedding space approximates similarity in the network.
Let’s define u and v as two nodes in a graph.
xu and xv are two feature vectors.
Now we’ll define the encoder function Enc(u) and Enc(v), which convert the feature vectors to zu and zv.
Note: the similarity function could be Euclidean distance.
So the challenge now is how to come up with the encoder function?
The encoder function should be able to perform :
- Locality (local network neighborhoods)
- Aggregate information
- Stacking multiple layers (computation)
Locality information can be achieved by using a computational graph. As shown in the graph below, i is the red node where we see how this node is connected to its neighbors and those neighbors’ neighbors. We’ll see all the possible connections, and form a computation graph.
By doing this, we’re capturing the structure, and also borrowing feature information at the same time.
Neighborhood exploration and information sharing | Source
Once the locality information preserves the computational graph, we start aggregating. This is basically done using neural networks.
Neural Networks are presented in grey boxes. They require aggregations to be order-invariant, like sum, average, maximum, because they are permutation-invariant functions. This property enables the aggregations to be performed.
Let’s move on to the forward propagation rule in GNNs. It determines how the information from the input will go to the output side of the neural network.
Deep Model: Many Layers | Source
Every node has a feature vector.
For example, (XA) is a feature vector of node A.
The inputs are those feature vectors, and the box will take the two feature vectors (XA and Xc), aggregate them, and then pass on to the next layer.
Notice that, for example, the input at node C are the features of node C, but the representation of node C in layer 1 will be a hidden, latent representation of the node, and in layer 2 it’ll be another latent representation.
So in order to perform forward propagation in this computational graph, we need 3 steps:
1. Initialize the activation units:
2. Every layer in the network:
We can notice that there are two parts for this equation:
- The first part is basically averaging all the neighbors of node v.
- The second part is the previous layer embedding of node v multiplied with a bias Bk, which is a trainable weight matrix and it’s basically a self-loop activation for node v.
- σ: the non-linearity activation that is performed on the two parts.
3. The last equation (at the final layer):
It’s the embedding after K layers of neighborhood aggregation.
Now, to train the model we need to define a loss function on the embeddings.
We can feed the embeddings into any loss function and run stochastic gradient descent to train the weight parameters.
Training can be unsupervised or supervised:
- Unsupervised training:
Use only the graph structure: similar nodes have similar embeddings. Unsupervised loss function can be a loss based on node proximity in the graph, or random walks.
- Supervised training:
Train model for a supervised task like node classification, normal or anomalous node.
To recap, in this section we described a basic idea of generating node embeddings by aggregating neighborhood information.
Next, I’ll discuss Graph Convolutional Networks (GCNs).
Graph Convolutional Networks
GCNs were first introduced in “Spectral Networks and Deep Locally Connected Networks on Graphs” (Bruna et al, 2014), as a method for applying neural networks to graph-structured data.
The simplest GCN has only three different operators:
- Graph convolution
- Linear layer
- Nonlinear activation
The operations are usually done in this order. Together, they make up one network layer. We can combine one or more layers to form a complete GCN.
In Python, we can easily build a GCN using PyTorch:
import torch from torch import nn class GCN(nn.Module): def __init__(self, *sizes): super().__init__() self.layers = nn.ModuleList([ nn.Linear(x, y) for x, y in zip(sizes[:-1], sizes[1:]) ]) def forward(self, vertices, edges): adj = torch.eye(len(vertices)) adj[edges[:, 0], edges[:, 1]] = 1 adj[edges[:, 1], edges[:, 0]] = 1 for layer in self.layers: vertices = torch.sigmoid(layer(adj @ vertices)) return vertices
GraphSAGE (Hamilton et al, NIPS 2017) is a representation learning technique for dynamic graphs.
It can predict the embedding of a new node, without needing a re-training procedure.
To do this, GraphSAGE uses inductive learning. It learns aggregator functions which can induce new node embedding, based on the features and neighborhood of the node.
We can notice two big differences. Rather than summing two things together and losing track of them, we use a general aggregation function which keeps them separate by concatenating them.
Before, we were using the Mean aggregation function – we simply took the message from the neighbors and added them up, and then normalized that by the number of neighbors. Now, we can also make a pooling type approach, or we can also use a deep neural network like an LSTM.
Neighbor Aggregation: Variants | Source
Applications of GNNs
As promised in the intro, I want to finish up with a few applications of GNNs.
Graph-structured data is present everywhere. The problems that GNNs resolve can be classified into these categories:
- Node Classification: the task here is to determine the labeling of samples (represented as nodes) by looking at the labels of their neighbors. Usually, problems of this type are trained in a semi-supervised way, with only a part of the graph being labeled.
- Graph Classification: the task here is to classify the whole graph into different categories. It’s like image classification, but the target changes into the graph domain. The applications of graph classification are numerous and range from determining whether a protein is an enzyme or not in bioinformatics, to categorizing documents in NLP, or social network analysis.
- Graph visualization: is an area of mathematics and computer science, at the intersection of geometric graph theory and information visualization. It is concerned with the visual representation of graphs that reveals structures and anomalies that may be present in the data and helps the user to understand the graphs.
- Link prediction: here, the algorithm has to understand the relationship between entities in graphs, and it also tries to predict whether there’s a connection between two entities. It’s essential in social networks to infer social interactions or to suggest possible friends to the users. It has also been used in recommender system problems and in predicting criminal associations.
- Graph clustering: refers to the clustering of data in the form of graphs. There are two distinct forms of clustering performed on graph data. Vertex clustering seeks to cluster the nodes of the graph into groups of densely connected regions based on either edge weights or edge distances. The second form of graph clustering treats the graphs as the objects to be clustered and clusters these objects based on similarity.
Let’s go through some applications across domains where GNN can resolve various challenges.
GNNs in computer vision
Using regular CNNs, machines can distinguish and identify objects in images and videos. Although there is still much development needed for machines to have the visual intuition of a human. Yet, GNN architectures can be applied to image classification problems.
One of these problems is scene graph generation, in which the model aims to parse an image into a semantic graph that consists of objects and their semantic relationships. Given an image, scene graph generation models detect and recognize objects and predict semantic relationships between pairs of objects.
However, the number of applications of GNNs in computer vision is still growing. It includes human-object interaction, few-shot image classification, and more.
GNNs in Natural Language Processing
In NLP, we know that the text is a type of sequential data which can be described by an RNN or an LSTM. However, graphs are heavily used in various NLP tasks, due to their naturalness and ease of representation.
Recently, there has been a surge of interest in applying GNNs for a large number of NLP problems like text classification, exploiting semantics in machine translation, user geolocation, relation extraction, or question answering.
We know that every node is an entity and edges describe relations between them. In NLP research, the problem of question answering is not recent. But it was limited by the existing database. Although, with techniques like GraphSage (Hamilton et al.), the methods can be generalized to previously unseen nodes.
GNNs in traffic
Forecasting traffic speed, volume or the density of roads in traffic networks is fundamentally important in a smart transportation system. We can address the traffic prediction problem by using STGNNs.
Considering the traffic network as a spatial-temporal graph where the nodes are sensors installed on roads, the edges are measured by the distance between pairs of nodes, and each node has the average traffic speed within a window as dynamic input features.
GNNs in chemistry
Chemists can use GNNs to research the graph structure of molecules or compounds. In these graphs, nodes are atoms, and edges – chemical bonds.
GNNs in other domains
The application of GNNs is not limited to the above domains and tasks. There have been attempts to apply GNNs to a variety of problems such as program verification, program reasoning, social influence prediction, recommender systems, electrical health records modeling, brain networks, and adversarial attack prevention.
Over the past few years, graph neural networks have become powerful and practical tools for any problem that can be modeled by graphs.
In this article, we did a comprehensive overview of graph neural networks and introduced a wide range of GNN applications.
If you stayed with me until the end – thank you for reading!
ML Experiment Tracking: What It Is, Why It Matters, and How to Implement It
Jakub Czakon | Posted November 26, 2020
Let me share a story that I’ve heard too many times.
”… We were developing an ML model with my team, we ran a lot of experiments and got promising results…
…unfortunately, we couldn’t tell exactly what performed best because we forgot to save some model parameters and dataset versions…
…after a few weeks, we weren’t even sure what we have actually tried and we needed to re-run pretty much everything”
– unfortunate ML researcher.
And the truth is, when you develop ML models you will run a lot of experiments.
Those experiments may:
- use different models and model hyperparameters
- use different training or evaluation data,
- run different code (including this small change that you wanted to test quickly)
- run the same code in a different environment (not knowing which PyTorch or Tensorflow version was installed)
And as a result, they can produce completely different evaluation metrics.
Keeping track of all that information can very quickly become really hard. Especially if you want to organize and compare those experiments and feel confident that you know which setup produced the best result.
This is where ML experiment tracking comes in.
Continue reading ->