and here is the magic - when you need build a data model of some spread - you have to create useful benchmarks, for a comparison of your results. Otherwise you will just shoot in the dark. Then I found similar regions in Italy based on:
- population
- area
- density
- GDP (economy data for better understanding of the population mobility)
- population in cities/villages
- number of similar cities and population there
- split from regions to counties
- distance from heavy impacted regions
- later discovered infected population than from patient zero
- and many more
These 4 regions have just 6% of Italy total infected population, but it's 12,6x more than in my country. First infected was discovered 26D (Marche) later from first in Italy. When we would like to "catch" their performance:
- we would have to grow from 6% to 35.6% per day for 8 consecutive days from April 6th, to get some numbers in Day 39. Which seems rather impossible
But when you open eye about daily increment of new infected persons, you will see a descent trend(except really small Molise region).