In other interesting news, a paper released this month describes a way of 'speeding up' neural net training, and an approach that achieves 4.9% top 5 validation error on Imagenet. My layperson's understanding is that this is the first time human accuracy has been exceeded on the Imagenet benchmarking challenge, and represents an advance on Chinese giant Baidu's progress reported last month, which I understood to be significant in its own right. http://arxiv.org/abs/1501.02876
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Sergey Ioffe, Christian Szegedy
(Submitted on 11 Feb 2015 (v1), last revised 13 Feb 2015 (this version, v2))
Training Deep Neural Networks is complicated by the fact that the distribution of each layer's inputs changes during training, as the parameters of the previous layers change. This slows down the training by requiring lower learning rates and careful parameter initialization, and makes it notoriously hard to train models with saturating nonlinearities. We refer to this phenomenon as internal covariate shift, and address the problem by normalizing layer inputs. Our method draws its strength from making normalization a part of the model architecture and performing the normalization for each training mini-batch. Batch Normalization allows us to use much higher learning rates and be less careful about initialization. It also acts as a regularizer, in some cases eliminating the need for Dropout. Applied to a state-of-the-art image classification model, Batch Normalization achieves the same accuracy with 14 times fewer training steps, and beats the original model by a significant margin. Using an ensemble of batch-normalized networks, we improve upon the best published result on ImageNet classification: reaching 4.9% top-5 validation error (and 4.8% test error), exceeding the accuracy of human raters."
Comment author:JWonz
08 March 2015 05:57:18PM
0 points
[-]
FYI - to those who are running the code, the atari ROMs must be named properly otherwise you will hit a segmentation fault. For example, with Breakout name it "breakout.bin".
= 27d567bc2d48d9f40e9ccc20ea370b15)
= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
= f037147d6e6c911a85753b9abdedda8d)
They've also released their code (for non-commercial purposes): https://sites.google.com/a/deepmind.com/dqn/
In other interesting news, a paper released this month describes a way of 'speeding up' neural net training, and an approach that achieves 4.9% top 5 validation error on Imagenet. My layperson's understanding is that this is the first time human accuracy has been exceeded on the Imagenet benchmarking challenge, and represents an advance on Chinese giant Baidu's progress reported last month, which I understood to be significant in its own right. http://arxiv.org/abs/1501.02876
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift Sergey Ioffe, Christian Szegedy
(Submitted on 11 Feb 2015 (v1), last revised 13 Feb 2015 (this version, v2))
Training Deep Neural Networks is complicated by the fact that the distribution of each layer's inputs changes during training, as the parameters of the previous layers change. This slows down the training by requiring lower learning rates and careful parameter initialization, and makes it notoriously hard to train models with saturating nonlinearities. We refer to this phenomenon as internal covariate shift, and address the problem by normalizing layer inputs. Our method draws its strength from making normalization a part of the model architecture and performing the normalization for each training mini-batch. Batch Normalization allows us to use much higher learning rates and be less careful about initialization. It also acts as a regularizer, in some cases eliminating the need for Dropout. Applied to a state-of-the-art image classification model, Batch Normalization achieves the same accuracy with 14 times fewer training steps, and beats the original model by a significant margin. Using an ensemble of batch-normalized networks, we improve upon the best published result on ImageNet classification: reaching 4.9% top-5 validation error (and 4.8% test error), exceeding the accuracy of human raters."
FYI - to those who are running the code, the atari ROMs must be named properly otherwise you will hit a segmentation fault. For example, with Breakout name it "breakout.bin".