Operational Calibration: Debugging Confidence Errors for DNNs in the Field

Cited by: 0|Bibtex|Views6|Links

Abstract:

Trained DNN models are increasingly adopted as integral parts of software systems. However, they are often over-confident, especially in practical operation domains where slight divergence from their training data almost always exists. To minimize the loss due to inaccurate confidence, operational calibration, i.e., calibrating the conf...More

Code:

Data:

Your rating :
0

 

Tags
Comments