Open LuminousPath opened 12 years ago
Productivity is a probability measure: double x, such that 0 <= x <= 1. If we are generating this value based on the productivity of two parent individuals from a "nursery" firm, then an acceptable standard algorithm must be established. --- CP: We'll use a bitset to represent an individual's ability to perform productively in a given Firm. The Hamming distance (HD) from skillSet to product will determine productivity within the Firm. Each individual will also be born with a random productivity coefficient (PC) that's generated by a normal distribution function. Thus, for a given Firm, an individual's relative productivity will be the PC*(HD(skillSet, product)/LENGTH(product)).
Are we basing the newly generated individual on other individuals in the firm or the overall firm itself?
Think of it like 2x mentor objects -> 1 new individual kerplop, birth, wahhhhh!
do we want them to hold general productivity counts, or different counts for different product ID's?
Individuals will probably be trickled into the system in order to keep population ages aligned with a normal distribution. --- CP: In order to accomplish this, we'll set the starting population's ages using a normal dist function, then have a normal dist function retirement rate.
Should have productivity, lifespan, age