for a given corpus are the Latent Dirichlet Allocation clusters for it is unique in general?
How about the gensim multi-process implementation of LDA? are there unique or they will be different for every run of code?
$\begingroup$
$\endgroup$
1
-
$\begingroup$ I'm not sure enough to write an answer, but as far as I remember there can be some randomness in the LDA process so potentially the clusters could be different. However if the data is large enough the differences are very small. But I'm not sure, so hopefully somebody will write an answer. $\endgroup$– ErwanNov 24, 2020 at 23:17
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
No, as there is randomness in the method implementation, for example here (in LdaModel of the gensim library). Hence, it can affect your final result in each run. Therefore, if you want to keep the result reproducible, you can set the random_state property of the model to a constant seed (see the documentation for more details).
