compare models

The front end.
Post Reply
cody
Posts: 6
Joined: Mon Apr 20, 2020 6:04 pm

compare models

Post by cody »

In using GeNIe, how can we compare two or three models and know which one is a better one? Literature typically indicates using sensitivity analysis and prediction accuracy. But is there any model fit index we can use to compare which model is a better one?

Relatedly, when we do structure learning, it produces a score. But when we revise this model, it does not produce any score. So we cannot compare scores between two models.

Thanks for any guidance and advice.

Cody
shooltz[BayesFusion]
Site Admin
Posts: 1417
Joined: Mon Nov 26, 2007 5:51 pm

Re: compare models

Post by shooltz[BayesFusion] »

If you have a dataset which can be used for prediction, use validation and check for accuracy or other parts of the confusion matrix. Use Learning|Validate command to access this functionality.
cody
Posts: 6
Joined: Mon Apr 20, 2020 6:04 pm

clustering algorithm

Post by cody »

I have another question. For clustering algorithm, I can select one algorithm in network. But I am confused about is that in the initial structure learning, I cannot specify which algorithm to use. It is only after the structure is learned, I can then specify which algorithm. Then what is the use of it since I have already had the structure learned? Please help me to understand its use. Thank you.
shooltz[BayesFusion]
Site Admin
Posts: 1417
Joined: Mon Nov 26, 2007 5:51 pm

Re: compare models

Post by shooltz[BayesFusion] »

The clustering algorithm is the inference algorithm (used to calculate the posterior probabilities). It is applicable once you have the network (both structure and parameters). Its inputs are the network itself and the evidence set in the network.

The structure learning algorithms use the data set as input. Their output is the structure of the network (nodes + arcs). Once the structure is available, the learning proceeds to parameter learning to obtain the CPTs.
Post Reply