Non-binary Decision Trees

The multiple must be made, not by always adding a higher dimension, but rather in the simplest of ways . . . write with slogans: Make rhizomes, not roots, never plant! Don't sow, grow offshoots! Don't be one or multiple, be multiplicities!

Deleuze & Guattari, A Thousand Plateaus (1987), pp. 6 . . . 24

A binary tree is a data structure that can be represented as a "tree" composed of branching paths between nodes. Each of these nodes presents an either-or option and the current path splits into left and right branches. This data structure can be quite powerful for driving algorithms and solving certain problems.

Binary tree (Wikipedia)

I require a more expansive approach so I've been developing a pen-and-paper technique for decision support that starts with a single path along a binary tree before expanding into the rest of the binary tree then breaking the structure altogether to support a broader set of possibilities. Each stage engages the question "where's the possibility of more possibilities?"

The exercise starts with a vertical line that has horizontal tick marks at each end and another in the middle. Beside each of these is a question. The bottom question is the most fundamental, the middle question is derived from the first and more specific, and the top question is derived from the middle and most specific. From this basis expansion begins.

I've identified 7 modes of expansion to date, 3 of which I'm willing to share publicly. These are:

• Vertical internal expansion of the line by adding more detail/options between existing bounds.
• Vertical external expansion of the line by adding more detail/options beyond existing bounds.
• Horizontal expansion by extending into new nodes on a broader decision tree.

Focused presence

My services

Home