A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is
We present an active learning strategy based on query by committee and dropout technique to train a Convolutional Neural Network (CNN). ... We propose a low computation adaptation of Query-By-Committee strategy (QBC) for deep learning. ... A simple strategy to extend active learning scheme to querying batch of unlabeled data, is to select the top scoring instances as already been applied for a previous deep active learning strategy  ...dblp:conf/esann/DucoffeP17 fatcat:uptfdozcezhr7hojc5nvsdxwk4
In this paper, we present an active learning strategy based on query by committee and dropout technique to train a Convolutional Neural Network (CNN). ... We evaluate our active learning strategy for CNN on MNIST benchmark, showing in particular that selecting less than 30 % from the annotated database is enough to get similar error rate as using the full ... ACKNOWLEDGMENTS We would like to thank the developpers of the frameworks Theano Bastien et al. (2012) , Blocks and Fuel van Merriënboer et al. (2015) which we used for the experi-ments. ...arXiv:1511.06412v2 fatcat:5rywqalfnfcnbafbz4px55ig3y
And of course Tilly, who is also the best and by far the coolest flatmate, thank you for all the memorable times we shared in our little crib. ... Thank you for being my rock and kick start in most difficult tasks, and being a dear friend to me. Thank you for showing me the way and the incredible kindness you bear. ... Acknowledgements Acknowledgements videos and learning their scripts by heart with me. ...doi:10.5075/epfl-thesis-9936 fatcat:4a7ntkikbngdvn47cy2ve2q7za