Machine Learning Workbench Propels TigerGraph-powered Machine Learning into the Limelight
- Blog >
- Machine Learning Workbench Propels TigerGraph-powered Machine Learning into the Limelight
When it was announced in May 2022, the TigerGraph Machine Learning Workbench promised to become an essential tool for data scientists who needed to harness the power of graph machine learning in an accessible, efficient way.
According to Dr. Victor Lee, Vice President of Machine Learning at TigerGraph, early adopters of the Machine Learning Workbench reported accuracy gains ranging from 10-50% in their machine learning models using the product.
That power and accessibility became available to all customers in the recent general release of the Machine Learning Workbench.
Low Effort and Familiar
In an interview for the TigerGraph Connections podcast, Victor explained the origin of the Machine Learning Workbench and how our customers would benefit from it: “We developed the Machine Learning Workbench to make it easier for data scientists to take advantage of graph in a very low effort, familiar way.”
He went on to explain that the product is built with a Python interface, an important inclusion since so many data scientists are very familiar with using Python in their day-to-day work. That familiarity allows them to issue commands that will use graph data and graph features to train the appropriate machine learning models for the project in question.
What does this all mean in practical terms for the organizations using TigerGraph’s Machine Learning Workbench? In short, they’ll benefit from deeper insights into their stores of data, improving the efficiency of use cases like fraud detection and product recommendations, to name just a couple.
Scaled to Your Needs
The TigerGraph Machine Learning Workbench can scale to work with enterprise-sized datasets but you don’t need to start there with the product. You can download the free developer edition of the software and have access to nearly all the functionality of the paid enterprise edition, with a few limitations, to start your machine learning journey with TigerGraph.
While you’ll have the same compatibility with TigerGraph databases, installation options, and built-in capabilities, the free version limits you to:
- 2GB data export
- No support for distributed parallel training
- Export via HTTP only
- Community support only
If you want more information about enterprise edition pricing, contact us for details.
Explore Machine Learning and Graph Neural Networks
Machine learning and graph neural networks have the ability to change business operations for organizations that can use them effectively. That can be a challenge, one summed up by Timo Klimmer, AI/ML Cloud Solution Architect at Microsoft:
“The main challenge in machine learning is usually not the algorithms. More often, it’s about feeding ML models with the right information, and graphs can be a good means to provide exactly that right information. Running TigerGraph’s Machine Learning Workbench on Azure helps our customers to leverage the power of graphs and combine it with the power of Azure Machine Learning.”
If you are looking to explore machine learning and graph neural networks in your organization, TigerGraph’s Machine Learning Workbench delivers a powerful combination of graph capabilities in an accessible, integrated package.
Learn more about TigerGraph’s Machine Learning Workbench here.