Comments (2)
Thank you so much for your response. That helps a lot!
from bnp.
Hi,
ASFAIK there is no best way to decide number of topics. HDP just assume the topics are infinite with DP prior (instead of a fix number K) so there is no guarantee it will find the right number of topics for you or the topics you see are consistent across different setting.
In practice, my personal experience is:
- when you increase
n_topic_truncate
, a topic will be split to multiple smaller and similar topics. (This is the same as LDA.) - how you initialize the stick break process also plays an important role on number of topics. You can try beta distribution which I commented out before. (https://github.com/chyikwei/bnp/blob/master/bnp/online_hdp.py#L439-L449)
And for n_topic_truncate=2
case, I tested it with dummy data and the result looks ok (I got 48%, 52% for topic proportion). You might need to turn hyper-parameters for your dataset.
>>> tf = make_doc_word_matrix(n_topics=5,
words_per_topic=10,
docs_per_topic=100,
words_per_doc=20,
shuffle=True,
random_state=0)
>>> hdp = HierarchicalDirichletProcess(n_topic_truncate=2, n_doc_truncate=3, max_iter=5, random_state=0)
>>> hdp.fit(tf)
HierarchicalDirichletProcess(alpha=1.0, batch_size=256, burn_in_iters=3,
check_doc_likelihood=False, eta=0.01, evaluate_every=0,
kappa=0.5, learning_method='online',
max_doc_update_iter=100, max_iter=5, mean_change_tol=0.001,
n_doc_truncate=3, n_jobs=1, n_topic_truncate=2, omega=1.0,
perp_tol=0.1, random_state=0, scale=1.0, tau=10.0,
total_samples=1000000, verbose=0)
>>> print_top_words(hdp, 10)
Topic 0 (proportion: 0.48): 3 1 7 5 19 10 4 8 0 2
Topic 1 (proportion: 0.52): 48 46 35 31 39 42 30 16 33 45
from bnp.
Related Issues (2)
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from bnp.