Database Administrators Stack Exchange is a question and answer site for database professionals who wish to improve their database skills and learn from others in the community. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm using SQL Server Analysis Studio and I'm trying to train a naive Bayes classifier. I've got an attribute called "item description" that basically represents products you can buy at a store. I've got these products mapped to a code and I want to train a classifier to take products it hasn't seen before and give me the code of highest probability.

What I'm running into is that my attribute of "item description" has too many states and so SSAS is throwing away a lot of my items - Analysis Studio says so. I'm got a case table that has an ID and code and a nested table that represents these "item descriptions" chopped up into individual words. Naive Bayes then uses conditional probabilities to map the words to a code.

I tinkered with the MAXIMUM_STATES setting and set it to 0 but according to the documentation, it maxes out at 65k.

Does anyone know if it's possible to raise this limit some magical way? Or can anyone suggest to me an alternate implementation of a naive Bayes classifier that is more scaleable.


migration rejected from Dec 11 '13 at 21:51

This question came from our site for professional and enthusiast programmers. Votes, comments, and answers are locked due to the question being closed here, but it may be eligible for editing and reopening on the site where it originated.

closed as off-topic by Shawn Melton, Paul White, bluefeet, Max Vernon, Michael - sqlbot Dec 11 '13 at 21:51

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "Too localized - this could be because your code has a typo, basic error, or is not relevant to most of our audience. Consider revising your question so that it appeals to a broader audience. As it stands, the question is unlikely to help other users (regarding typo questions, see this meta question for background)." – Shawn Melton, Paul White, bluefeet, Max Vernon, Michael - sqlbot
If this question can be reworded to fit the rules in the help center, please edit the question.

You could always implement it yourself... Naive Bayes is so simple that it fits into one or two queries. The formulae are also quite simple (you sum the log-odds of all features for each document). – usr Aug 14 '13 at 16:27
Could you provide us a snapshot displaying part of the model as shown at the SSAS' Naive Bayes viewer? – Marcus Vinicius Pompeu Oct 4 '13 at 2:57
Sorry, it was so long ago I don't have it anymore. Lately I've started looking at Mahout to do the same thing and see how it holds up. – mj_ Oct 4 '13 at 14:37