search

ymli39 / DeepSEED-3D-ConvNets-for-Pulmonary-Nodule-Detection

DeepSEED: 3D Squeeze-and-Excitation Encoder-Decoder ConvNets for Pulmonary Nodule Detection
MIT License
109 stars 33 forks source link

The prediction probability is not in 0-1 range? #21

Closed KinaTech closed 3 years ago

KinaTech commented 3 years ago

Hi, thanks for the wonderful explanation. When I test your model, I found the prediction probability is not in 0-1 range. I mean the probability is like 4.5, 18, 21.5. Does these phenomenon normal? It looks like these are the diameter of the nodules. But how can I get the probabilities? I attach an example of output of the model. Thanks in advance @ymli39 

   [ -2.97817516, 285.57787803, 285.55532278, 425.68683523,
      9.22237027],
   [ -2.78340006, 285.48476638, 285.41791546, 429.51884933,
      5.54941623],
   [ -2.89723754, 285.46616549, 285.3996337 , 429.56445795,
      9.20982807]])
kingjames1155 commented 1 year ago

Hi, thanks for the wonderful explanation. When I test your model, I found the prediction probability is not in 0-1 range. I mean the probability is like 4.5, 18, 21.5. Does these phenomenon normal? It looks like these are the diameter of the nodules. But how can I get the probabilities? I attach an example of output of the model. Thanks in advance @ymli39

   [ -2.97817516, 285.57787803, 285.55532278, 425.68683523,
      9.22237027],
   [ -2.78340006, 285.48476638, 285.41791546, 429.51884933,
      5.54941623],
   [ -2.89723754, 285.46616549, 285.3996337 , 429.56445795,
      9.20982807]])

I have encountered the same problem. Have you solved it