Using graphs can improve machine learning and produce explainable AI. Boost your machine learning arsenal, learn how graph algorithms apply to unsupervised learning, and build machine learning pipelines that integrate graph analytics with conventional machine learning.
Remember: Our main program features more knowledge graph & AI greatness from the likes of Microsoft, Uber, nVidia, Bayer, GSK, and JP Morgan, combo tickets on sale!
Need to convince your manager? We got you a kit, start working on it!
This workshop is kindly sponsored by TigerGraph. TigerGraph is the fastest and most scalable graph database analytics platform—
Machine Learning, Graph Algorithms
Data scientists, application developers, or graph users who want to know some practical ways that graphs can enhance machine learning and explainable AI.
Intermediate – Advanced
Participants should be familiar with basic concepts of both machine learning and property graphs, but they do not need to be proficient developers. Participants should have past experience writing and running simple snippets of code.
Bring your own laptop. A modern web browser required. Note: some of the applications do not support Internet Explorer.
● Different ways that graph data and graph analytics can improve machine learning and produce explainable AI.
● Major types of graph algorithms and how they apply to unsupervised learning.
● Graph features and a simple example of a graph query to extract a feature from all the nodes in a graph.
● Outline a machine learning pipeline that integrates graph analytics with conventional machine learning.
● Draw the connection between graph visualization and explainable AI.
Graphs are changing the rules for machine learning. Traditionally, data must be structured as a 2D matrix or 3D tensor for analysis, and feature dimensionality reduction is often used.
Graphs, on the other hand, are an intuitive, adaptable, and efficient way to represent knowledge and relationships in an unlimited number of dimensions, where each type of relationship represents an additional dimension. In short, graphs provide a potentially richer source of knowledge, but data scientists need to know practical methods for leveraging this resource.
The course combines lecture, demonstration, and hands-on exercises to introduce to how scalable graph analytics platforms are being used today for machine learning and explainable AI.
We start by looking at different types of machine learning and the stages of a pipeline, and consider where graph structure and analytics can add value. We will consider what it means to “do machine learning in the graph”.
We will do a hands-on survey of graph algorithms, to see how they can directly support unsupervised learning.
We will next see how some algorithms, and graph queries in general, are a powerful technique for feature extraction.
We will look at the specific case of graph embedding, transforming a graph into a set of vectors.
We will then introduce a common open-source machine learning platform and see how to exchange data between a graph database and the ML environment.
The course will culminate by implementing a full pipeline: graph data preparation, graph feature extraction, exporting from graph database to ML platform, training, testing, and deploying the model for prediction.
The course will also give an introduction to more advanced techniques and areas of open research for graph-integrative machine learning.
We built our workshops based on a set of simple principles, to make sure you get the most out of them: