event thumbnail image
ICML 2007 - The 24th Annual International Conference on Machine Learning
PASCAL

Statistical Predicate Invention

author: Stanley Kok, University of Washington

Description

We propose statistical predicate invention as a key problem for statistical relational learning. SPI is the problem of discovering new concepts, properties and relations in structured data, and generalizes hidden variable discovery in statistical models and predicate invention in ILP. We propose an initial model for SPI based on second-order Markov logic, in which predicates as well as arguments can be variables, and the domain of discourse is not fully known in advance. Our approach iteratively refines clusters of symbols based on the clusters of symbols they appear in atoms with (e.g., it clusters relations by the clusters of the ob jects they relate). Since different clusterings are better for predicting different subsets of the atoms, we allow multiple cross-cutting clusterings. We show that this approach outperforms Markov logic structure learning and the recently introduced infinite relational model on a number of relational datasets.

You might be experiencing some problems with Your Video player.
Slides
0:00 Statistical Predicate Invention
0:07 - Motivation
0:24 Motivation - 1
0:51 Motivation - 2
1:33 SPI Benefits
2:25 State of the Art
3:12 Multiple Relational Clusterings - 1
3:50 - Background
3:52 Markov Logic Networks (MLNs) - 1
4:27 Markov Logic Networks (MLNs) - 2
5:01 - Multiple Relational Clusterings
5:04 Multiple Relational Clusterings - 2
5:36 Example of Multiple Clusterings
6:15 Second-Order Markov Logic
6:53 Symbols
7:41 MRC Rules - 1
8:16 MRC Rules - 2
9:18 Learning MRC Model - 1
9:51 Learning MRC Model - 2
10:21 Learning MRC Model - 3
11:04 Search Algorithm - 1
12:05 Search Algorithm - 2
12:09 Methodology - 1
12:27 Methodology - 2
13:09 Results
15:16 Multiple Clusterings Learned - 1
15:52 Multiple Clusterings Learned - 2
16:07 Multiple Clusterings Learned - 3
16:11 - Future Work
16:11 Future Work
16:41 - Questions

Lecture rating

People found this lecture:
Worth seeing
because it is:
 Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.

Related content

Link this page

Would you like to put a link to this lecture on your homepage?
Go ahead! Copy the HTML snippet !

Write your own review or comment: