Skip to main content

New answers tagged

0 votes

My semantic segmentation model classifies everything as background

Class weights are probably the best first choice when fighting foreground/background imbalance. But another similar approach is just not computing the loss/metrics over the background. If you do that, ...
Luan Souza's user avatar
0 votes

Keras LSTM model not performant

You could try several different things here. Since it seems like overfitting might be an issue here, you could try early stopping call-backs: ...
healthydata's user avatar
0 votes

Validation accuracy can't increase above 70%

It seems like the net is confusing 1 with 2, and 2 with 1. This suggests the net is not able to distinguish between the classes. Some domain knowledge could help with understanding why they get mixed ...
MuhammedYunus's user avatar

Top 50 recent answers are included