this is for holding javascript data
Richard Guo added In_general_this_condition_does__.tex
over 8 years ago
Commit id: 63fa89afa1e10b9f02544f441b64278519944655
deletions | additions
diff --git a/In_general_this_condition_does__.tex b/In_general_this_condition_does__.tex
new file mode 100644
index 0000000..ae7e990
--- /dev/null
+++ b/In_general_this_condition_does__.tex
...
In general, this condition does not hold. Therefore, we could model a hashtag series as an observed sum of two (or more) interacting Hawkes processes. For every event $i$ generated by $N(t)$, we could introduce a latent variable $z_i \in \{1,2\}$ that indicates whether it comes from $N_1(t)$ or $N_2(t)$. In particular, since we also have the user and RT (retweet) information in the data, and we could assume that the latent variable is determined by the author (or RT) information (stochastic block model). Equivalently, from a bottom-up perspective, this amounts to grouping users who post with a hashtag into two (or more) opposing parties and from the model we could map the standing of each user (the probability of being assigned to cluster 1 vs cluster 2).
