blob: 5e54737e1d076ba30f7225b5826dfc1c8a9448d8 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
|
can find p_s ourselves — don’t need to restrict to earlier papers
cutoff at three steps vs no cutoff — compare for appendix
how would epidemiologists approach this
demographics — risk factors for infection
similarity — e.g. exposure to people diff age from you
gang membership as binary, also if in same gang
propensity score for beta values?
model vs model with demographics?
goal — don’t get laughed at. just get into review and revision
try for hybrid model
that could be a separate paper
could show more possible infections, or people who could have been infected by multiple people so involved in a cascade
how accurate are cascades? — show scores
can cascades predict what happened
understanding cascades can help save x lives
what percent of shootings were result of cascades
can we distinguish which ones are not
distribution cascade/individuals over time or given certain traits
add in fatal shootings — just so it looks the same in the network
can compare which are more likely to cause infections
all we’re looking for is exposure so someone who is shot and killed can still infect a friend
try multiple vs single infections, for people who have been shot multiple times. compare results
|
