Timeline for Sampling posterior of empty cluster in GMM and Gibbs
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 19, 2018 at 4:29 | comment | added | rano | I see what you mean, but the extreme example is somehow outside the contemplated cases : ) Freezing the allocations might be interesting on the other hand, see @Xi'an answer below | |
| Jul 17, 2018 at 23:03 | comment | added | guy | The point was to give an example to show why it is much better to update them. The mixing of the chain (assuming it actually corresponds to some stationary distribution) will potentially be made much worse by freezing; I was just giving an extreme example. I cannot see any upside in freezing, aside from saving a pretty negligible amount of compute time. | |
| Jul 17, 2018 at 20:41 | comment | added | rano | actually I was saying to freezing them to the last configuration where the cluster had been assigned some instances | |
| Jul 16, 2018 at 20:35 | comment | added | guy | Irrespective of whether it is valid (it is not), it is a bad idea to freeze. Imagine you initialize in a way that makes one cluster have really bad parameters; because of this, the cluster is empty. Now you are stuck with these bad initial values effectively forever because you are not changing them. If you update them using the prior, you should eventually by chance have them be reasonable values, at which point you might get a new cluster appearing. | |
| Jul 16, 2018 at 20:29 | answer | added | Xi'an | timeline score: 1 | |
| May 28, 2018 at 13:52 | history | asked | rano | CC BY-SA 4.0 |