Presentation is loading. Please wait.

Presentation is loading. Please wait.

Pegna, J.M., Lozano, J.A., and Larragnaga, P.

Similar presentations


Presentation on theme: "Pegna, J.M., Lozano, J.A., and Larragnaga, P."— Presentation transcript:

1 Learning recursive Bayesian multinets for data clustering by means of constructive induction
Pegna, J.M., Lozano, J.A., and Larragnaga, P. Machine Learning, 47(1), pp , 2002. Summarized by Kyu-Baek Hwang

2 data clustering problem
data partitioning k-means representing the joint probability distribution of a database mixture density models (e.g., Gaussian mixtures) Bayesian networks C Y1 Y2 Y3 Y4 Y5 (c) 2003 SNU CSE Biointelligence Laboratory

3 recursive Bayesian multinets
RBMN a decision tree where each decision path ends in an alternate component Bayesian network (BN) context-specific conditional independencies (c) 2003 SNU CSE Biointelligence Laboratory

4 BNs for data clustering
the joint probability distribution by a BN (c) 2003 SNU CSE Biointelligence Laboratory

5 (c) 2003 SNU CSE Biointelligence Laboratory
Bayesian multinets encode the context-specific conditional independencies. (c) 2003 SNU CSE Biointelligence Laboratory

6 (c) 2003 SNU CSE Biointelligence Laboratory
RBMNs extensions of BMNs or partitional clustering systems (c) 2003 SNU CSE Biointelligence Laboratory

7 (c) 2003 SNU CSE Biointelligence Laboratory
real world domain geographical distribution of malignant tumors (c) 2003 SNU CSE Biointelligence Laboratory

8 component BN structures
extended naïve Bayes (ENB) models a selection of the attributes to be included in the models (X) some attributes can be grouped together under the same node (O) (c) 2003 SNU CSE Biointelligence Laboratory

9 learning algorithm for ENB models
(c) 2003 SNU CSE Biointelligence Laboratory

10 (c) 2003 SNU CSE Biointelligence Laboratory
parameter search EM (expectation maximization) algorithm BC (bound and collapse) + EM algorithm (O) (c) 2003 SNU CSE Biointelligence Laboratory

11 (c) 2003 SNU CSE Biointelligence Laboratory
structure search constructive induction the process of changing the representation of the cases in the database by creating new attributes from existing attributes. forward algorithm backward algorithm (c) 2003 SNU CSE Biointelligence Laboratory

12 marginal likelihood criterion for RBMNs
with uninformative Dirichlet prior, for BMNs, with some reasonable assumptions including parameter independence, for RBMNs, (c) 2003 SNU CSE Biointelligence Laboratory

13 learning algorithm for RBMNs
(c) 2003 SNU CSE Biointelligence Laboratory

14 (c) 2003 SNU CSE Biointelligence Laboratory
experimental setup both synthetic data and real data discrete variables with (unrestricted) multinomial distributions convergence criterion for BC + EM algorithm change in the log marginal likelihood value is less than 10-6 or 150 iterations fixing_probability_threshold: 0.51 initial structure: naïve Bayes model 5 independent runs at each experiment (c) 2003 SNU CSE Biointelligence Laboratory

15 1-level RBMNs for the experiments
(c) 2003 SNU CSE Biointelligence Laboratory

16 2-level RBMNs for the experiments
(c) 2003 SNU CSE Biointelligence Laboratory

17 performance for 4 synthetic databases
(c) 2003 SNU CSE Biointelligence Laboratory

18 performance for real world data
tic-tac-toe data: 2 clusters with 9 predictive variables, 958 cases nursery data: 5 clusters with 8 predictive variables, cases (c) 2003 SNU CSE Biointelligence Laboratory

19 conclusions and future research
context-specific conditional independencies data partitioning efficient representation, Bayesian committees, mixture of experts learning speed problem trade-off with the efficient representation monothetic decision tree polythetic paths  enrich the modeling power extensions to the continuous domain (c) 2003 SNU CSE Biointelligence Laboratory


Download ppt "Pegna, J.M., Lozano, J.A., and Larragnaga, P."

Similar presentations


Ads by Google