Slow Treebagger Prediction Time for One Datapoint
4 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
Hi,
I'm looking to use a Treebagger model as part of a project, and I'd like to call this model in a loop, with inference time under 0.1 seconds. I've trained a few models on data, and I've noticed that calling predict(Model,Examples(1)) takes ~2 seconds, while calling predict(Model,Examples(1:1000)) takes a similar amount of time, and predict(Model,Examples(1:10000)) takes ~3 seconds. This seems to mean that there's a startup time associated with calling predict(Model,Examples()).
Is there a way to use this Treebagger model in a loop, and avoid the ~2 second startup time each time I call it?
Note: when testing this on individual trees from the Treebagger model, I've found a proportionally similar startup time for predict().
Any help is appreciated.
Best,
SP
0 comentarios
Respuestas (0)
Ver también
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!