Implementation

This section shows how to use an advanced version of naive Bayes and where to insert this component in the CRLMM described in Chapter 11, Conceptual Representation Learning.

Gaussian naive Bayes

In implementation mode, a dataset with raw data from the blockchain will be used without the feature interpretation function of naive Bayes in the following table:

DAY	STOCK	BLOCKS	DEMAND
10	1455	78	1
11	1666	67	1
12	1254	57	1
14	1563	45	1
15	1674	89	1
10	1465	89	1
12	1646	76	1
15	1746	87	2
12	1435	78	2

 

Each line represents a block:

• DAY: The day of the month scanned (dd/mm/yyyy can be used beyond a prototype)
• STOCK: The total inputs in a given location (A, B, or... F) found in the blocks and totaled on that day
• BLOCKS: The number of blocks containing product P for location A, for example

A high number of blocks in the BLOCK column and a low number of quantities in the STOCK column mean that the demand is high. 

• DEMAND = 1. The proof of demand is a transaction block that contains a purchase in the past. These transaction blocks...