The (a, q) data modeling in probabilistic reasoning

This article considers a critical and experimental approach on the attributive and qualitative AI data modeling and data retrieval in computational probabilistic reasoning. The mathematical correlation of X≈Y in the d=dx/dy differentiations and its point based locations and matrix based predictions in Markov Models, Rete’s Algorithms and Bayesian fields, with the further development of non-linear ‘human-type’ reasoning in AI. The new approach in the ternary system transition (decimal↔binary) of Brusentsov-Bergman principle by its bound allocation in the ‘mirror-based’ system in t... t powers, and hereon considers its further data retrieval for suitable matching and translation of probabilistic data differentiation. The causation/probability matrix of this paper regards not only bound/free variable in x1,x2,x3, x variables, but discovers and explains its further subsets in aXq formula, where the supposition of d=X/Y regarded not as a mathematical placement of the variable X, but as its attributive (a) and qualitative (q) allocation in a certain value/relevance cell of the Probability Triangle of the ternary system. From where the automated differentiation retrieves only the most relevant/objective aXq data cell, not the closest by the pre-set context, making the AI selections more assertive and preference based than linear.