Sunday, November 13, 2016

Is it me or is network theory getting big in economics?

Probablyme, I keep looking to see who has new goodies.

Let me chimer in a bit on heory.
The general approach is to define a network structure,then imply maximum entropy flow from one region to another. If there is a common value line, a universal red/green; then the disordering of that flow, relative to the optimum Huffman flow, the disordering is the cost of  the pit boss, I have an example.

The mathematician grocer who runs the local corner store.  He has two queues, the gods in and the goods out.  They are not nearly equal. He has a segmented customer base, and he knits them together by maintaining a segmented supply network. He has customers thatshop at the Grocer outlet, less often; and drop by to get small things.  So, his wife is off to grocer outlet, buying buy the case o make up for their fractional bit error.  He does the floating point part of multiply.  That lets Grocer Outlet, buy and sell in big cases, most of the time, and their trucks are mostly full.

He does this for about three small sequences of schoolgirls customers. This might include running by Safemart for three gallons of mils on the way to open the store, thusstabilizing the flow of milk lovers, a segmented group, actual;y.

The disorder is real, real in the sense it meets Ito's condition, this grocer is doing standard continuous flow accounting, without much error.  His accounting is accurate because he, on net, meets the Shannon condition in his subgraph, he wont; be requantized; he keeps the big guys efficient.  It is like he forced himself on heir payroll by going outside the gates and putting order into the mob..


No comments: