Posted Content•
Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization.
TL;DR: The results suggest that regularization is important for worst-group generalization in the overparameterized regime, even if it is not needed for average generalization, and introduce a stochastic optimization algorithm, with convergence guarantees, to efficiently train group DRO models.
Abstract: Overparameterized neural networks can be highly accurate on average on an i.i.d. test set yet consistently fail on atypical groups of the data (e.g., by learning spurious correlations that hold on average but not in such groups). Distributionally robust optimization (DRO) allows us to learn models that instead minimize the worst-case training loss over a set of pre-defined groups. However, we find that naively applying group DRO to overparameterized neural networks fails: these models can perfectly fit the training data, and any model with vanishing average training loss also already has vanishing worst-case training loss. Instead, the poor worst-case performance arises from poor generalization on some groups. By coupling group DRO models with increased regularization---a stronger-than-typical L2 penalty or early stopping---we achieve substantially higher worst-group accuracies, with 10-40 percentage point improvements on a natural language inference task and two image tasks, while maintaining high average accuracies. Our results suggest that regularization is important for worst-group generalization in the overparameterized regime, even if it is not needed for average generalization. Finally, we introduce a stochastic optimization algorithm, with convergence guarantees, to efficiently train group DRO models.
Citations
More filters
Posted Content•
TL;DR: WILDS is presented, a benchmark of in-the-wild distribution shifts spanning diverse data modalities and applications, and is hoped to encourage the development of general-purpose methods that are anchored to real-world distribution shifts and that work well across different applications and problem settings.
Abstract: Distribution shifts -- where the training distribution differs from the test distribution -- can substantially degrade the accuracy of machine learning (ML) systems deployed in the wild. Despite their ubiquity, these real-world distribution shifts are under-represented in the datasets widely used in the ML community today. To address this gap, we present WILDS, a curated collection of 8 benchmark datasets that reflect a diverse range of distribution shifts which naturally arise in real-world applications, such as shifts across hospitals for tumor identification; across camera traps for wildlife monitoring; and across time and location in satellite imaging and poverty mapping. On each dataset, we show that standard training results in substantially lower out-of-distribution than in-distribution performance, and that this gap remains even with models trained by existing methods for handling distribution shifts. This underscores the need for new training methods that produce models which are more robust to the types of distribution shifts that arise in practice. To facilitate method development, we provide an open-source package that automates dataset loading, contains default model architectures and hyperparameters, and standardizes evaluations. Code and leaderboards are available at this https URL.
579 citations
Cites background or methods from "Distributionally Robust Neural Netw..."
...We note that reweighting methods like Group DRO are effective only when the training loss is non-vanishing, which we achieve through early stopping (Byrd & Lipton, 2019; Sagawa et al., 2020a;b)....
[...]
...…al., 2018a), ImageNet-C (Hendrycks & Dietterich, 2019), and similar ImageNet variants (Geirhos et al., 2018b); and datasets that crop out objects and replace their backgrounds, as in the Backgrounds Challenge (Xiao et al., 2020) and other similar datasets (Sagawa et al., 2020a; Koh et al., 2020)....
[...]
...We adapted the implementations of CORAL from Gulrajani & Lopez-Paz (2020); IRM from Arjovsky et al. (2019); and Group DRO from Sagawa et al. (2020a)....
[...]
...We fixed the step size hyperparameter for Group DRO to its default value of 0.01 (Sagawa et al., 2020a)....
[...]
...These types of spurious correlations can significantly degrade model performance on particular subpopulations (Sagawa et al., 2020a)....
[...]
Posted Content•
TL;DR: This paper implements DomainBed, a testbed for domain generalization including seven multi-domain datasets, nine baseline algorithms, and three model selection criteria, and finds that, when carefully implemented, empirical risk minimization shows state-of-the-art performance across all datasets.
Abstract: The goal of domain generalization algorithms is to predict well on distributions different from those seen during training While a myriad of domain generalization algorithms exist, inconsistencies in experimental conditions -- datasets, architectures, and model selection criteria -- render fair and realistic comparisons difficult In this paper, we are interested in understanding how useful domain generalization algorithms are in realistic settings As a first step, we realize that model selection is non-trivial for domain generalization tasks Contrary to prior work, we argue that domain generalization algorithms without a model selection strategy should be regarded as incomplete Next, we implement DomainBed, a testbed for domain generalization including seven multi-domain datasets, nine baseline algorithms, and three model selection criteria We conduct extensive experiments using DomainBed and find that, when carefully implemented, empirical risk minimization shows state-of-the-art performance across all datasets Looking forward, we hope that the release of DomainBed, along with contributions from fellow researchers, will streamline reproducible and rigorous research in domain generalization
492 citations
Cites methods from "Distributionally Robust Neural Netw..."
...For example, to implement group DRO [Sagawa et al., 2019, Algorithm 1], we simply write the following in algorithms.py: class DRO(ERM): def __init__(self , input_shape , num_classes , num_domains , hparams ): super ()....
[...]
Posted Content•
TL;DR: This work introduces the principle of Risk Extrapolation (REx), and shows conceptually how this principle enables extrapolation, and demonstrates the effectiveness and scalability of instantiations of REx on various OoD generalization tasks.
Abstract: Distributional shift is one of the major obstacles when transferring machine learning prediction systems from the lab to the real world. To tackle this problem, we assume that variation across training domains is representative of the variation we might encounter at test time, but also that shifts at test time may be more extreme in magnitude. In particular, we show that reducing differences in risk across training domains can reduce a model's sensitivity to a wide range of extreme distributional shifts, including the challenging setting where the input contains both causal and anti-causal elements. We motivate this approach, Risk Extrapolation (REx), as a form of robust optimization over a perturbation set of extrapolated domains (MM-REx), and propose a penalty on the variance of training risks (V-REx) as a simpler variant. We prove that variants of REx can recover the causal mechanisms of the targets, while also providing some robustness to changes in the input distribution ("covariate shift"). By appropriately trading-off robustness to causally induced distributional shifts and covariate shift, REx is able to outperform alternative methods such as Invariant Risk Minimization in situations where these types of shift co-occur.
400 citations
Posted Content•
TL;DR: On x-ray grading and bird identification, concept bottleneck models achieve competitive accuracy with standard end-to-end models, while enabling interpretation in terms of high-level clinical concepts (“bone spurs”) or bird attributes ( “wing color”).
Abstract: We seek to learn models that we can interact with using high-level concepts: if the model did not think there was a bone spur in the x-ray, would it still predict severe arthritis? State-of-the-art models today do not typically support the manipulation of concepts like "the existence of bone spurs", as they are trained end-to-end to go directly from raw input (e.g., pixels) to output (e.g., arthritis severity). We revisit the classic idea of first predicting concepts that are provided at training time, and then using these concepts to predict the label. By construction, we can intervene on these concept bottleneck models by editing their predicted concept values and propagating these changes to the final prediction. On x-ray grading and bird identification, concept bottleneck models achieve competitive accuracy with standard end-to-end models, while enabling interpretation in terms of high-level clinical concepts ("bone spurs") or bird attributes ("wing color"). These models also allow for richer human-model interaction: accuracy improves significantly if we can correct model mistakes on concepts at test time.
226 citations
Posted Content•
TL;DR: It is found that there is often little to no transfer of robustness from current synthetic to natural distribution shift, and the results indicate that distribution shifts arising in real data are currently an open research problem.
Abstract: We study how robust current ImageNet models are to distribution shifts arising from natural variations in datasets. Most research on robustness focuses on synthetic image perturbations (noise, simulated weather artifacts, adversarial examples, etc.), which leaves open how robustness on synthetic distribution shift relates to distribution shift arising in real data. Informed by an evaluation of 204 ImageNet models in 213 different test conditions, we find that there is often little to no transfer of robustness from current synthetic to natural distribution shift. Moreover, most current techniques provide no robustness to the natural distribution shifts in our testbed. The main exception is training on larger and more diverse datasets, which in multiple cases increases robustness, but is still far from closing the performance gaps. Our results indicate that distribution shifts arising in real data are currently an open research problem. We provide our testbed and data as a resource for future work at this https URL .
212 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
Journal Article•
TL;DR: It is shown that dropout improves the performance of neural networks on supervised learning tasks in vision, speech recognition, document classification and computational biology, obtaining state-of-the-art results on many benchmark data sets.
Abstract: Deep neural nets with a large number of parameters are very powerful machine learning systems. However, overfitting is a serious problem in such networks. Large networks are also slow to use, making it difficult to deal with overfitting by combining the predictions of many different large neural nets at test time. Dropout is a technique for addressing this problem. The key idea is to randomly drop units (along with their connections) from the neural network during training. This prevents units from co-adapting too much. During training, dropout samples from an exponential number of different "thinned" networks. At test time, it is easy to approximate the effect of averaging the predictions of all these thinned networks by simply using a single unthinned network that has smaller weights. This significantly reduces overfitting and gives major improvements over other regularization methods. We show that dropout improves the performance of neural networks on supervised learning tasks in vision, speech recognition, document classification and computational biology, obtaining state-of-the-art results on many benchmark data sets.
33,597 citations
Book•
[...]
TL;DR: In this article, the focus is on recognizing convex optimization problems and then finding the most appropriate technique for solving them, and a comprehensive introduction to the subject is given. But the focus of this book is not on the optimization problem itself, but on the problem of finding the appropriate technique to solve it.
Abstract: Convex optimization problems arise frequently in many different fields. A comprehensive introduction to the subject, this book shows in detail how such problems can be solved numerically with great efficiency. The focus is on recognizing convex optimization problems and then finding the most appropriate technique for solving them. The text contains many worked examples and homework exercises and will appeal to students, researchers and practitioners in fields such as engineering, computer science, mathematics, statistics, finance, and economics.
33,341 citations
Proceedings Article•
06 Jul 2015TL;DR: 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.
Abstract: 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, and in some cases eliminates 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 4.82% top-5 test error, exceeding the accuracy of human raters.
30,843 citations
"Distributionally Robust Neural Netw..." refers methods in this paper
...As in the original paper, we used batch normalization (Ioffe & Szegedy, 2015) and no dropout (Srivastava et al., 2014)....
[...]
11 Oct 2018
TL;DR: BERT as mentioned in this paper pre-trains deep bidirectional representations from unlabeled text by jointly conditioning on both left and right context in all layers, which can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of tasks.
Abstract: We introduce a new language representation model called BERT, which stands for Bidirectional Encoder Representations from Transformers. Unlike recent language representation models (Peters et al., 2018a; Radford et al., 2018), BERT is designed to pre-train deep bidirectional representations from unlabeled text by jointly conditioning on both left and right context in all layers. As a result, the pre-trained BERT model can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of tasks, such as question answering and language inference, without substantial task-specific architecture modifications. BERT is conceptually simple and empirically powerful. It obtains new state-of-the-art results on eleven natural language processing tasks, including pushing the GLUE score to 80.5 (7.7 point absolute improvement), MultiNLI accuracy to 86.7% (4.6% absolute improvement), SQuAD v1.1 question answering Test F1 to 93.2 (1.5 point absolute improvement) and SQuAD v2.0 Test F1 to 83.1 (5.1 point absolute improvement).
24,672 citations