The classic 3*4+1+1+4+(9-1)/2+[this one @23rd]+(9-1)/2. For a total of 27. The whole 163 and x^2-x+41 Technetium (+2) connection. Interesting things in number theory along with sporadic groups and J4 which is the only one with an ordered factor of 43 and an 11^3. Promethium at 61 is connected somehow maybe by 12 * 62 = 744 with something not doing the 10 “f-orbitals” thing, and 23 comes in on the uniqueness of factorization too along with 105.  Along with the 18 families of groups 26(or 27)+18 = 44(or 45) in cubic elliptic varieties of the discriminant.

26 letters in the alphabet plus space? Rocks with patterned circles on an island? Considering one of the 44 is the circle integer modulo ring with no “torsion” then there is kind of 43 bending varieties and some kind of dimension null over a double bend “cover” inclusion as a half factor of one of the main 18 sequence groups. Likely a deep connection to factor square-free “Mobius mu” and topological orientability.

Polynomial Regression Estimators

Consider a sampled sequence of n samples and an interpolation of order n. The sample sequence can be differentiated by backward and forward differences of all n samples to make a first differential sequence of n elements or more. This too has a polynomial fit. The polynomial can be integrated to make an order n+1 polynomial with a new constant which can be estimated by a regression fit of the n samples. This can then make an n+1 th estimation to show a fit ad infinitum. Weighting the regression error based on sample time locks more history and less prediction into the forecast but fits less on the predictive end. Opposite the forecast is based on a forecast not based on history. In between is a concept of optimal.

A genetic algorithm optimizing the weighting provides a fitness score based on future measured truth. The population spread acts as a Monte-Carlo and some selection for spreading entropy as well as future weight would input entropy flair for efficiency by the association of prediction clustering elimination and outlier promotion for risk estimates. An irony of population size and death by accounting in genetic algorithms weeds out some ”bum notes’ ‘ but “right on” in the ill computed silicon heaven (via Lobb’s theorem of truth by confirmed assumption). Hence an eviction cache as in silicon hardware. What measures the crash instability of markets in the recession local optimum?

Yes, I do imply logic machines are operating reality. I do not think all the machines use the same operator algebra. Some algebras survive, some do not. There is nothing in the closure complexity of efficient algebras supporting the accumulation of axioms as leisure free from a suppressed fight.