Calculus of Thought

Chapter 4

Causal Reasoning

Abstract

Reduced error logistic regression's (RELR) causal reasoning mechanism that is based upon outcome score matching was reviewed with a focus on methodology issues that have plagued propensity score matching methods. Like all other RELR artificial intelligence processes, this process can be implemented as a completely automatic data-driven computational mechanism. The essence of this process is a matched sample quasi-experimental process to test hypothesized causal effects that have been initially learned through Explicit RELR. This process uses Implicit RELR to produce a very high-dimensional offset regression model that computes the effect of all other factors on the outcome probability while holding the putative ...