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
TL;DR: NAS-PERIRB as mentioned in this paper uses the early stopping and insert replay buffer (IRB) to improve the training efficiency of the samples and design the Pareto efficiency reward function to optimize the goals.
6 citations
17 Oct 2022
TL;DR: A novel neural architecture search framework for automated spatio-temporal synchronous modeling in traffic prediction, entitled AutoSTS, is proposed, where multiple graphs built from different perspectives are jointly utilized to find a better message passing way for mining such correlations.
Abstract: Traffic prediction plays an important role in many intelligent transportation systems. Many existing works design static neural network architecture to capture complex spatio-temporal correlations, which is hard to adapt to different datasets. Although recent neural architecture search approaches have addressed this problem, it still adopts a coarse-grained search with pre-defined and fixed components in the search space for spatio-temporal modeling. In this paper, we propose a novel neural architecture search framework, entitled AutoSTS, for automated spatio-temporal synchronous modeling in traffic prediction. To be specific, we design a graph neural network (GNN) based architecture search module to capture localized spatio-temporal correlations, where multiple graphs built from different perspectives are jointly utilized to find a better message passing way for mining such correlations. Further, we propose a convolutional neural network (CNN) based architecture search module to capture temporal dependencies with various ranges, where gated temporal convolutions with different kernel sizes and convolution types are designed in search space. Extensive experiments on six public datasets demonstrate that our model can achieve 4%-10% improvements compared with other methods.
6 citations
05 Dec 2021
TL;DR: In this article, a trust-region strategy is proposed to minimize the time-to-market of analog design space search, which is better cast as constraint satisfaction problem than global optimization defined in prior arts.
Abstract: This paper introduces new perspectives on analog design space search. To minimize the time-to-market, this endeavor better cast as constraint satisfaction problem than global optimization defined in prior arts. We incorporate model based agents, contrasted with model-free learning, to implement a trust-region strategy. As such, simple feed-forward networks can be trained with supervised learning, where the convergence is relatively trivial. Experiment results demonstrate orders of magnitude improvement on search iterations. Additionally, the unprecedented consideration of PVT conditions are accommodated. On circuits with TSMC 5/6nm process, our method achieve performance surpassing human designers. Furthermore, this framework is in production in industrial settings.
6 citations
TL;DR: A novel search method conducted directly on camera trap images with lowered resolutions and maintained aspect ratios is designed, guided by a loss function whose hyper parameter is theoretically derived for finding lightweight networks.
Abstract: Simple Summary Camera traps acquire visual data in a non-disturbing and round-the-clock manner, so they are popular for ecological researchers observing wildlife. Each camera trap may record thousands of images of diverse species and bring about millions of images that need to be classified. Many methods have been proposed to classify camera trap images, but almost all methods rely on very deep convolutional neural networks that require intensive computational resources. Such resources may be unavailable and become formidable in cases where the surveillance area is large or becomes greatly expanded. We turn our attention to camera traps organized as groups, where each group produces images that are processed by the edge device with lightweight networks tailored for images produced by the group. To achieve this goal, we propose a method to automatically design networks deployable for edge devices with respect to given images. With the proposed method, researchers without any experience in designing neural networks can develop networks applicable for edge devices. Thus, camera trap images can be processed in a distributed manner through edge devices, lowering the costs of transferring and processing data accumulated at camera traps. Abstract Camera traps provide a feasible way for ecological researchers to observe wildlife, and they often produce millions of images of diverse species requiring classification. This classification can be automated via edge devices installed with convolutional neural networks, but networks may need to be customized per device because edge devices are highly heterogeneous and resource-limited. This can be addressed by a neural architecture search capable of automatically designing networks. However, search methods are usually developed based on benchmark datasets differing widely from camera trap images in many aspects including data distributions and aspect ratios. Therefore, we designed a novel search method conducted directly on camera trap images with lowered resolutions and maintained aspect ratios; the search is guided by a loss function whose hyper parameter is theoretically derived for finding lightweight networks. The search was applied to two datasets and led to lightweight networks tested on an edge device named NVIDIA Jetson X2. The resulting accuracies were competitive in comparison. Conclusively, researchers without knowledge of designing networks can obtain networks optimized for edge devices and thus establish or expand surveillance areas in a cost-effective way.
6 citations
TL;DR: The ability of all three approaches to predict when individual vehicles could potentially connect to charging stations to be comparable, resulting in the same set of 30 vehicles identified as good candidates to participate in a vehicle-to-grid service is found.
Abstract: Vehicle-to-grid services draw power or curtail demand from electric vehicles when they are connected to a compatible charging station. In this paper, we investigated automated machine learning for predicting when vehicles are likely to make such a connection. Using historical data collected from a vehicle tracking service, we assessed the technique’s ability to learn and predict when a fleet of 48 vehicles was parked close to charging stations and compared this with two moving average techniques. We found the ability of all three approaches to predict when individual vehicles could potentially connect to charging stations to be comparable, resulting in the same set of 30 vehicles identified as good candidates to participate in a vehicle-to-grid service. We concluded that this was due to the relatively small feature set and that machine learning techniques were likely to outperform averaging techniques for more complex feature sets. We also explored the ability of the approaches to predict total vehicle availability and found that automated machine learning achieved the best performance with an accuracy of 91.4%. Such technology would be of value to vehicle-to-grid aggregation services.
6 citations
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)....
[...]