September 2021 Summaries

The blog post discusses a method for executing decision trees inside a TigerGraph instance, which is useful in scenarios where explainable machine learning models are important. The author uses a banking dataset to train the decision tree and then loads the data and decision tree into the TigerGraph instance. They utilize accumulators in TigerGraph to make classifications by traversing the decision tree within the graph structure. This approach allows users to run interpretable ML models inside of TigerGraph, where their data resides, and can be applied to various business processes such as claims adjudication. The code for this implementation is available on GitHub.