Closed Bokyeong1001 closed 1 year ago
It's hard to be able to pin point what might be the source of the problem w/o having more information on the dataset..
In our experience DSM typically works atleast as well if not better than Deephit on most datasets we experimented with.
Have you tried tuning the DSMs temperature parameter? 'temp' perhaps between [1, 50, 100] ? (Default is 1000).
If that doesn't help, feel free to email me at
@Bokyeong1001 i think there might be a bug in your code.
You aren't really using the model with the lowest nll, you're reusing the last model. You need to make sure you're using the model with the lowest neg log likelihood.
Thank you so much!
I will fix the bug and try again with tuning 'temp' parameter.
If it doesn't work, I will email with information about dataset.
Hello!
I am trying to apply Deep Survival Machine to my dataset which has 23 events. And the results are not as good as I expected.
In the DSM paper, with SEER data, DSM and DeepHit c-index results are pretty comparable, but with my data DeepHit c-index results are about 10% better than DSM.
I made code based on your DSM example notebook code. I only changed the evaluation part since the example notebook code is for a single event.
Is it possible that DSM is unsuitable for data with many events? Or just because my code has a problem?
Here is my code.
`
`