Back to all updates

about 1 year ago

TigerGraph Machine Learning Workbench

We’re excited to announce that TigerGraph Machine Learning Workbench is now available!

Video: TigerGraph Machine Learning Workbench Overview

TigerGraph ML Workbench is a Jupiter-based Python framework that allows you to develop AI and Machine Learning models on top of a TigerGraph solution. With TigerGraph ML Workbench, you can now easily explore the potentials of Graph Neural Networks for your domains!


Success Stories of Graph Neural Networks

GNN has proven its success both in academic and industrial settings. It tends to outperform other traditional machine learning techniques when there are well-defined relationships between data as it directly models the connectivities of graph data. Listed below are some references on how GNN models are transforming a wide range of applications and industries to spark new ideas on how much more we can do for our customers.

  • Recommendation Engine
    Pinterest introduced PinSAGE [1], an architecture that can serve real-time recommendations to their users, resulting in a 10-30% improvement compared to other deep learning methods when evaluated in A/B testing.

  • Supply Chain
    Amazon released a GNN architecture [2] that incorporates temporal information with GNNs for demand forecasting. The method models interactions between products and sellers on Amazon in a graph, resulting in a 16% improvement over other state-of-the-art forecasting methods.

  • Healthcare
    AstraZeneca has used graph neural networks like GraphSAGE to generate knowledge graph embeddings for predicting possible drug-drug interactions such as potential synergies between drugs, as well as possible polypharmacy side effects [3]. Additionally, the possibility of repurposing drugs to treat COVID has been studied using a drug repurposing knowledge graph and GNNs [4].

  • Financial Institutions
    GCNs have been studied for predicting money-laundering behavior in Bitcoin transaction networks and have been shown to perform admirably compared to other approaches [5].


TigerGraph Machine Learning Workbench Quick Start Instructions

Whether you are new to TigerGraph or are a current user, we have a path to get you started quickly.

  • (Option 1) ML Workbench Sandbox - This Docker image includes the TigerGraph database with data set preloaded, ML Workbench preconfigured, and example GNN notebooks written in python to kick start your GNN development. This is the quickest way to play around with ML Workbench.

  • (Option 2) ML Workbench Standalone - If you already have a TigerGraph database instance set up and want to develop GNNs with your data set, you can download this standalone image and connect ML Workbench to your TigerGraph server. This image also comes with example GNN notebooks written in python as a template recipe.

  • (Option 3) ML Workbench for existing Jupyter Server - If you already have a TigerGraph database instance set up AND an existing notebook server running on-prem or with a third-party cloud platform such as Amazon SageMaker, you can instead download our ML Workbench python kernel.


TG Cloud Account Reminder

Remember to sign up for your TG Cloud account to get started with your hackathon project. Use the below link to signup for your TG Cloud account with the email used for Devpost.

Happy Building!