Notes on the Behavior of MC Dropout

Research output: Contribution to conferencePaperScientificpeer-review


Among the various options to estimate uncertainty in deep neural networks, Monte-Carlo dropout is widely popular for its simplicity and effectiveness. However the quality of the uncertainty estimated through this method varies and choices in architecture design and in training procedures have to be carefully considered and tested to obtain satisfactory results. In this paper we present a study offering a different point of view on the behavior of Monte-Carlo dropout, which enables us to observe a few interesting properties of the technique to keep in mind when considering its use for uncertainty estimation.
Original languageEnglish
Number of pages6
Publication statusPublished - 23 Jul 2021
MoE publication typeNot Eligible
EventICML Workshop on Uncertainty & Robustness in Deep Learning - Virtual, Online
Duration: 23 Jul 202123 Jul 2021
Conference number: 2021


WorkshopICML Workshop on Uncertainty & Robustness in Deep Learning
Abbreviated titleICML UDL
CityVirtual, Online
Internet address


Dive into the research topics of 'Notes on the Behavior of MC Dropout'. Together they form a unique fingerprint.

Cite this