Last time I looked at these hidden Markov models. We will get the complete text book, good for two semesters of undergrad. A student can almost pull all the material from Wiki.
I want graphics, dogs, cats, proteins, protons, swirls of vacuum, pictures of toroidal space. All the neat stuff.
Graphic of surface matching in action fro the AI path minimizers. Dynamic graphics, show us N density, or, rather, regions of biased deviation counts, recolored regions that do not close, explosions.
A generic package that goes from Markov tree to surfaces, in n color; up the n tuples. Add your own kinetic rules of engagement, make them walk and talk.
The animation is great, you have a lot of freedom in describing the movement of centroids. Just using the 4-tuiple and managing the equations of motion by an artificial fourth color which really allows you to alter surface angles. The color opertor will still make whit. Relativity means that all your equations of motions can be modeled as movements and currents in vacuum density, leading automatically to 3D rendering and shading closed surfaces. A much better Pixar. this subject came up years ago on this blog, it is real.
Running the 4 tuple means you can maintain the vertical you want to measure. Shows the attachment vector in protein immuno globulins, distribution of vacuum around Mercury. Surface shape of a dog running.
But the method applies to all data that can be conceived of as matching surfaces along some dimension.
No comments:
Post a Comment