Using the Shapley's Additive Explanation Algorithm to Explain a Decision Tree

If I could use one tool to explain any model would be shop, or any tool that's based on Shapley's additive explanation algorithm. Unfortunately, that algorithm can't really be applied to modern ML because of the data sizes that we usually deal with in training data sets. For classification, you're going to determine like is this the same? What is the probability of class membership for these two? So if that row was predicted without manipulation as we belong into class number two, and then with manipulation, it's class zero or class one, what was the probability that it was actually of class two? And we look at that difference and you actually determine how far off or how close that prediction