Thoughts About Data Science

Data science is calculus. And we don't use the complete curve to understand it. So, having the complete curve can't help us in any way to make better predictions? That's silly. By using multiple functions, even hundreds, and combining them to define the whole curve, is what we should be doing in all things. Especially engineering and physics.

And if you say we define math, that's teleological. The structures that math describe are inherent in those designs. So, if we're talking about Physics, what could not having the entire curve do? Beside give us incomplete data? And if we're talking about mathematics, why would we need something, that doesn't actually model anything? Be it a code--which is a form of geometry--or a set, or what have you? Because although the geometry isn't connected together in physical space, it still is a geometric pattern. And that's what we have to understand. Because physicists do not get to play God with their data. It needs to have the whole thing interpreted.

All mathematics are rooted in geometry. Including the axioms we use. The axioms we use, move like logic, but if there's nothing real by which the axioms can be used for, then they're not truly axioms. They're just circular and teleological. Everything has to be rooted in reality. Even if it's hyperdimensional space, but we can't just use logic for the sake of logic. Because logic breaks down, when there's nothing real to base the logic on. It's just endless circles. The logic of algebra always works--and you can make true logical chains--but then logic begins to break down, when you're just using logic of logic of logic of logic. As where's the first principles or axioms to begin with in that matter? They're non existent. And there's nothing to test them against, to see if they're actually true. Where Calculus, we can test it, and see it actually works. And so can a quadratic equation. And so can a Pythagorean Theorem, or any of the other millions of applications, from physics to engineering.