Posted Content•
Neural Architecture Search with Reinforcement Learning
TL;DR: This paper uses a recurrent network to generate the model descriptions of neural networks and trains this RNN with reinforcement learning to maximize the expected accuracy of the generated architectures on a validation set.
Abstract: Neural networks are powerful and flexible models that work well for many difficult learning tasks in image, speech and natural language understanding. Despite their success, neural networks are still hard to design. In this paper, we use a recurrent network to generate the model descriptions of neural networks and train this RNN with reinforcement learning to maximize the expected accuracy of the generated architectures on a validation set. On the CIFAR-10 dataset, our method, starting from scratch, can design a novel network architecture that rivals the best human-invented architecture in terms of test set accuracy. Our CIFAR-10 model achieves a test error rate of 3.65, which is 0.09 percent better and 1.05x faster than the previous state-of-the-art model that used a similar architectural scheme. On the Penn Treebank dataset, our model can compose a novel recurrent cell that outperforms the widely-used LSTM cell, and other state-of-the-art baselines. Our cell achieves a test set perplexity of 62.4 on the Penn Treebank, which is 3.6 perplexity better than the previous state-of-the-art model. The cell can also be transferred to the character language modeling task on PTB and achieves a state-of-the-art perplexity of 1.214.
Citations
More filters
25 Apr 2022
TL;DR: This paper proposes a novel Multimodal Structure-evolving Continual Graph Learning (MSCGL) model, which continually learns both the model architecture and the corresponding parameters for Adaptive Multinodal Graph Neural Network (AdaMGNN).
Abstract: Continual graph learning is rapidly emerging as an important role in a variety of real-world applications such as online product recommendation systems and social media. While achieving great success, existing works on continual graph learning ignore the information from multiple modalities (e.g., visual and textual features) as well as the rich dynamic structural information hidden in the ever-changing graph data and evolving tasks. However, considering multimodal continual graph learning with evolving topological structures poses great challenges: i) it is unclear how to incorporate the multimodal information into continual graph learning and ii) it is nontrivial to design models that can capture the structure-evolving dynamics in continual graph learning. To tackle these challenges, in this paper we propose a novel Multimodal Structure-evolving Continual Graph Learning (MSCGL) model, which continually learns both the model architecture and the corresponding parameters for Adaptive Multimodal Graph Neural Network (AdaMGNN). To be concrete, our proposed MSCGL model simultaneously takes social information and multimodal information into account to build the multimodal graphs. In order for continually adapting to new tasks without forgetting the old ones, our MSCGL model explores a new strategy with joint optimization of Neural Architecture Search (NAS) and Group Sparse Regularization (GSR) across different tasks. These two parts interact with each other reciprocally, where NAS is expected to explore more promising architectures and GSR is in charge of preserving important information from the previous tasks. We conduct extensive experiments over two real-world multimodal continual graph scenarios to demonstrate the superiority of the proposed MSCGL model. Empirical experiments indicate that both the architectures and weight sharing across different tasks play important roles in affecting the model performances.
12 citations
TL;DR: By employing new methods from deep learning, higher performance in terms of accuracy and lower inference time can be achieved, and the models can be made useful in real-world applications.
Abstract: Deep Learning has been successfully applied to image recognition, speech recognition, and natural language processing in recent years. Therefore, there has been an incentive to apply it in other fields as well. The field of agriculture is one of the most important fields in which the application of deep learning still needs to be explored, as it has a direct impact on human well-being. In particular, there is a need to explore how deep learning models can be used as a tool for optimal planting, land use, yield improvement, production/disease/pest control, and other activities. The vast amount of data received from sensors in smart farms makes it possible to use deep learning as a model for decision-making in this field. In agriculture, no two environments are exactly alike, which makes testing, validating, and successfully implementing such technologies much more complex than in most other industries. This paper reviews some recent scientific developments in the field of deep learning that have been applied to agriculture, and highlights some challenges and potential solutions using deep learning algorithms in agriculture. The results in this paper indicate that by employing new methods from deep learning, higher performance in terms of accuracy and lower inference time can be achieved, and the models can be made useful in real-world applications. Finally, some opportunities for future research in this area are suggested.
12 citations
Journal Article•
TL;DR: A model-free estimator inspired by the marginal likelihood, which is used for Bayesian model selection, that can achieve a rank correlation of 0.95 with final test accuracy on the NAS-Bench201 dataset within 50 epochs is proposed.
Abstract: Reliable yet efficient evaluation of generalisation performance of a proposed architecture is crucial to the success of neural architecture search (NAS). Traditional approaches face a variety of limitations: training each architecture to completion is prohibitively expensive, early stopping estimates may correlate poorly with fully trained performance, and model-based estimators require large training sets. Instead, motivated by recent results linking training speed and generalisation with stochastic gradient descent, we propose to estimate the final test performance based on the sum of training losses. Our estimator is inspired by the marginal likelihood, which is used for Bayesian model selection. Our model-free estimator is simple, efficient, and cheap to implement, and does not require hyperparameter-tuning or surrogate training before deployment. We demonstrate empirically that our estimator consistently outperforms other baselines under various settings and can achieve a rank correlation of 0.95 with final test accuracy on the NAS-Bench201 dataset within 50 epochs.
12 citations
Cites background from "Neural Architecture Search with Rei..."
...The traditional approach of training each architecture for a large number of epochs and evaluating it on validation data (full evaluation) provides a reliable performance measure, but requires prohibitively high computational resources on the order of thousands of GPU days [2, 3, 4, 5, 1]....
[...]
Posted Content•
TL;DR: This dissertation focuses on improving the efficiency of deep neural networks at four levels, using automated neural architecture search algorithms to discover models with state-of-the-art accuracy and efficiency.
Abstract: Author(s): Wu, Bichen | Advisor(s): Keutzer, Kurt W | Abstract: The success of deep neural networks (DNNs) is attributable to three factors: increased compute capacity, more complex models, and more data. These factors, however, are not always present, especially for edge applications such as autonomous driving, augmented reality, and internet-of-things. Training DNNs requires a large amount of data, which is difficult to obtain. Edge devices such as mobile phones have limited compute capacity, and therefore, require specialized and efficient DNNs. However, due to the enormous design space and prohibitive training costs, designing efficient DNNs for different target devices is challenging. So the question is, with limited data, compute capacity, and model complexity, can we still successfully apply deep neural networks?This dissertation focuses on the above problems and improving the efficiency of deep neural networks at four levels. Model efficiency: we designed neural networks for various computer vision tasks and achieved more than 10x faster speed and lower energy. Data efficiency: we developed an advanced tool that enables 6.2x faster annotation of a LiDAR point cloud. We also leveraged domain adaptation to utilize simulated data, bypassing the need for real data. Hardware efficiency: we co-designed neural networks and hardware accelerators and achieved 11.6x faster inference. Design efficiency: the process of finding the optimal neural networks is time-consuming. Our automated neural architecture search algorithms discovered, using 421x lower computational cost than previous search methods, models with state-of-the-art accuracy and efficiency.
12 citations
Cites background from "Neural Architecture Search with Rei..."
...Starting from [187], recent works adopt neural architecture search (NAS) to explore the design space automatically....
[...]
...Many previous works [187, 188, 141] use reinforcement learning (RL) to guide the search and a typical flow is illustrated in Figure 8....
[...]
...Neural Architecture Search: [187, 188] first propose to use reinforcement learning (RL) to search for neural architectures to achieve competitive accuracy with low FLOPs....
[...]
...Most of the previous works [187, 188, 110, 92, 93] search for cell level architectures....
[...]
...Previous works [187, 188, 141] have demonstrated the effectiveness of such methods in finding accurate and efficient networks....
[...]
TL;DR: The LM with MMLSTMs surpasses the existing state-of-the-art model on Penn Treebank and WikiText-2 data sets and outperforms the baseline by 3.3 points in perplexity onWikiText-103 data set without increasing model parameter counts.
Abstract: Language model (LM) plays an important role in natural language processing (NLP) systems, such as machine translation, speech recognition, learning token embeddings, natural language generation, and text classification. Recently, the multilayer long short-term memory (LSTM) models have been demonstrated to achieve promising performance on word-level language modeling. For each LSTM layer, larger hidden size usually means more diverse semantic features, which enables the LM to perform better. However, we have observed that when a certain LSTM layer reaches a sufficiently large scale, the promotion of overall effect will slow down, as its hidden size increases. In this article, we analyze that an important factor leading to this phenomenon is the high correlation between the newly extended hidden states and the original hidden states, which hinders diverse feature expression of the LSTM. As a result, when the scale is large enough, simply lengthening the LSTM hidden states will cost tremendous extra parameters but has little effect. We propose a simple yet effective improvement on each LSTM layer consisting of a large-scale Major LSTM and a small-scale Minor LSTM to break the high correlation between the two parts of hidden states, which we call Major–Minor LSTMs (MMLSTMs). In experiments, we demonstrate the LM with MMLSTMs surpasses the existing state-of-the-art model on Penn Treebank (PTB) and WikiText-2 (WT2) data sets and outperforms the baseline by 3.3 points in perplexity on WikiText-103 data set without increasing model parameter counts.
12 citations
Cites background from "Neural Architecture Search with Rei..."
...But in fact, the Minor LSTM is essentially different from existing shortcut connections [27], [30], [31], [34], [35]....
[...]
References
More filters
27 Jun 2016
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.
123,388 citations
Proceedings Article•
01 Jan 2015TL;DR: This work introduces Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments, and provides a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework.
Abstract: We introduce Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments. The method is straightforward to implement, is computationally efficient, has little memory requirements, is invariant to diagonal rescaling of the gradients, and is well suited for problems that are large in terms of data and/or parameters. The method is also appropriate for non-stationary objectives and problems with very noisy and/or sparse gradients. The hyper-parameters have intuitive interpretations and typically require little tuning. Some connections to related algorithms, on which Adam was inspired, are discussed. We also analyze the theoretical convergence properties of the algorithm and provide a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework. Empirical results demonstrate that Adam works well in practice and compares favorably to other stochastic optimization methods. Finally, we discuss AdaMax, a variant of Adam based on the infinity norm.
111,197 citations
Proceedings Article•
04 Sep 2014TL;DR: This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.
Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.
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)....
[...]
01 Jan 1998
TL;DR: In this article, a graph transformer network (GTN) is proposed for handwritten character recognition, which can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters.
Abstract: Multilayer neural networks trained with the back-propagation algorithm constitute the best example of a successful gradient based learning technique. Given an appropriate network architecture, gradient-based learning algorithms can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters, with minimal preprocessing. This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task. Convolutional neural networks, which are specifically designed to deal with the variability of 2D shapes, are shown to outperform all other techniques. Real-life document recognition systems are composed of multiple modules including field extraction, segmentation recognition, and language modeling. A new learning paradigm, called graph transformer networks (GTN), allows such multimodule systems to be trained globally using gradient-based methods so as to minimize an overall performance measure. Two systems for online handwriting recognition are described. Experiments demonstrate the advantage of global training, and the flexibility of graph transformer networks. A graph transformer network for reading a bank cheque is also described. It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal cheques. It is deployed commercially and reads several million cheques per day.
42,067 citations
20 Jun 2005
TL;DR: It is shown experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection, and the influence of each stage of the computation on performance is studied.
Abstract: We study the question of feature sets for robust visual object recognition; adopting linear SVM based human detection as a test case. After reviewing existing edge and gradient based descriptors, we show experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection. We study the influence of each stage of the computation on performance, concluding that fine-scale gradients, fine orientation binning, relatively coarse spatial binning, and high-quality local contrast normalization in overlapping descriptor blocks are all important for good results. The new approach gives near-perfect separation on the original MIT pedestrian database, so we introduce a more challenging dataset containing over 1800 annotated human images with a large range of pose variations and backgrounds.
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)....
[...]