Neural Architecture Search with Reinforcement Learning
Citations
40 citations
Cites background or methods from "Neural Architecture Search with Rei..."
..., reinforcement learning [41, 42], evolution [20], gradient optimization [15, 17], Bayesian optimization [39]), and plays an important role in NAS [41, 42, 18, 20, 17, 15, 39]....
[...]
...For reinforcement learning based algorithms [41, 42, 18] where the controller is usually an RNN model, we can predict the accuracy of the architectures generated by the RNN and take the predicted accuracy as the reward to train the controller....
[...]
...Popular methods usually consume hundreds to thousands of GPU days to discover eventually good architectures [41, 20, 17]....
[...]
...Conventional NAS includes [41, 42, 20, 17], which achieve significant improvements on several benchmark datasets....
[...]
39 citations
Cites background or methods from "Neural Architecture Search with Rei..."
...Pioneers in this field develop prototypes based on reinforcement learning [43], evolutionary algorithms [30] and Bayesian optimization [24]....
[...]
...00 949 Early methods in NAS usually include a full training and evaluation procedure every iteration as the inner loop to guide the consecutive search [30, 43, 44]....
[...]
39 citations
39 citations
Cites background from "Neural Architecture Search with Rei..."
...In [41], an RNN-based policy network is proposed which iteratively generates new architectures based on the gradient information from its child networks....
[...]
39 citations
Cites background from "Neural Architecture Search with Rei..."
...Fueled by the promising progress of NAS, we propose to introduce NAS into the action recognition task with an efficient fashion....
[...]
...From this perspective, neural architecture search can be treated as a kind of meta-learning, in which NAS is to transfer the network trained on the training data to adapt the validation data by fine-tuning the architecture....
[...]
...Such promising progresses are mostly benefited from the current strong computational capability, the elaborate search space [13] as well as the efficiency of search strategies like reinforcement learning [7], evolutionary algorithms [9] and gradient methods [14]....
[...]
...Recently, Neural Architecture Search (NAS) [7] has shown great superiority over manually designed neural architectures....
[...]
...In image classification, automatically designed architectures like NASNet [8], and AmoebaNet [9], outperform popular human-designed networks, such as VGG [10], and ResNet [11]....
[...]
References
123,388 citations
111,197 citations
55,235 citations
"Neural Architecture Search with Rei..." refers methods in this paper
...Along with this success is a paradigm shift from feature designing to architecture designing, i.e., from SIFT (Lowe, 1999), and HOG (Dalal & Triggs, 2005), to AlexNet (Krizhevsky et al., 2012), VGGNet (Simonyan & Zisserman, 2014), GoogleNet (Szegedy et al., 2015), and ResNet (He et al., 2016a)....
[...]
42,067 citations
31,952 citations
"Neural Architecture Search with Rei..." refers methods in this paper
...Along with this success is a paradigm shift from feature designing to architecture designing, i.e., from SIFT (Lowe, 1999), and HOG (Dalal & Triggs, 2005), to AlexNet (Krizhevsky et al., 2012), VGGNet (Simonyan & Zisserman, 2014), GoogleNet (Szegedy et al., 2015), and ResNet (He et al., 2016a)....
[...]