外刊聽(tīng)力練習(xí) 經(jīng)濟(jì)學(xué)人 tracking the fighting with s

Russia’s invasion of Ukraine have devastated the country.
But the full scale of military activity has been hard to assess
So Sandra (人名) the Economist senior data journalist came up with a plan to penetrate
the fog of war
?
I figured my comparative advantage would be to look at the war from above.
One worry I had when the war began was that a lot of media outlets were relying on the same sources of information
it ’s pretty easy to see what’s going on in Kiev right now
You have a lot of people there
You have a lot of reporter there
?
It is slightly harder to see what is going on close to the front line on the Ukrainian side
It is way harder to know what’s going on on the Russian side
And especially deeply in Russian occupied areas
?
I wanted an independent system
A system that was neutral consistent
?And that could track the fighting without relying on human input
Firms is a satellite surveillance system set up By Nasa to track the forest fires
However it tracks a lot of other fires too
?
I set up a system that looks at this forest fire data and ?
?Tries to figure out if it is really forest fires or any other source of fire
That would happen in peace time or it is related to the fighting
?
I first said about constructing what we call training data
So this is data from normal years
That The model we use to sort of learn what is supposed to be going on
You need to go back in time and find all sorts of other data
But once you have that you go into the training stage
And that is telling a computer find all the patterns you can in this data
And in my case telling it to do that one hundred times
?
And then if you plot the abnormal Activity ?trends start to appear
This map reveals areas where the fires are active
If you look at where the fires are most concentrated we can see a hot spot in the road to Kiev
And the front line can also be clearly observed.
?
?