D. Sculley

Shreya Shankar, Yoni Halpern, Eric Breck et al.

Modern machine learning systems such as image classifiers rely heavily on large scale data sets for training. Such data sets are costly to create, thus in practice a small number of freely available, open source data sets are widely used. We suggest that examining the geo-diversity of open data sets is critical before adopting a data set for use cases in the developing world. We analyze two large, publicly available image data sets to assess geo-diversity and find that these data sets appear to exhibit an observable amerocentric and eurocentric representation bias. Further, we analyze classifiers trained on these data sets to assess the impact of these training distributions and find strong differences in the relative performance on images from different locales. These results emphasize the need to ensure geo-representation when constructing data sets for use in the developing world.

Daniel Smilkov, Shan Carter, D. Sculley et al.

The recent successes of deep learning have led to a wave of interest from non-experts. Gaining an understanding of this technology, however, is difficult. While the theory is important, it is also helpful for novices to develop an intuitive feel for the effect of different hyperparameters and structural variations. We describe TensorFlow Playground, an interactive, open sourced visualization that allows users to experiment via direct manipulation rather than coding, enabling them to quickly build an intuition about neural nets.

Alexander D'Amour, Katherine Heller, Dan Moldovan et al.

ML models often exhibit unexpectedly poor behavior when they are deployed in real-world domains. We identify underspecification as a key reason for these failures. An ML pipeline is underspecified when it can return many predictors with equivalently strong held-out performance in the training domain. Underspecification is common in modern ML pipelines, such as those based on deep learning. Predictors returned by underspecified pipelines are often treated as equivalent based on their training domain performance, but we show here that such predictors can behave very differently in deployment domains. This ambiguity can lead to instability and poor model behavior in practice, and is a distinct failure mode from previously identified issues arising from structural mismatch between training and deployment domains. We show that this problem appears in a wide variety of practical ML pipelines, using examples from computer vision, medical imaging, natural language processing, clinical risk prediction based on electronic health records, and medical genomics. Our results show the need to explicitly account for underspecification in modeling pipelines that are intended for real-world deployment in any domain.

Zachary Nado, Shreyas Padhy, D. Sculley et al.

Covariate shift has been shown to sharply degrade both predictive accuracy and the calibration of uncertainty estimates for deep learning models. This is worrying, because covariate shift is prevalent in a wide range of real world deployment settings. However, in this paper, we note that frequently there exists the potential to access small unlabeled batches of the shifted data just before prediction time. This interesting observation enables a simple but surprisingly effective method which we call prediction-time batch normalization, which significantly improves model accuracy and calibration under covariate shift. Using this one line code change, we achieve state-of-the-art on recent covariate shift benchmarks and an mCE of 60.28\% on the challenging ImageNet-C dataset; to our knowledge, this is the best result for any model that does not incorporate additional data augmentation or modification of the training pipeline. We show that prediction-time batch normalization provides complementary benefits to existing state-of-the-art approaches for improving robustness (e.g. deep ensembles) and combining the two further improves performance. Our findings are supported by detailed measurements of the effect of this strategy on model behavior across rigorous ablations on various dataset modalities. However, the method has mixed results when used alongside pre-training, and does not seem to perform as well under more natural types of dataset shift, and is therefore worthy of additional study. We include links to the data in our figures to improve reproducibility, including a Python notebooks that can be run to easily modify our analysis at https://colab.research.google.com/drive/11N0wDZnMQQuLrRwRoumDCrhSaIhkqjof.

Christof Angermueller, David Belanger, Andreea Gane et al.

The use of black-box optimization for the design of new biological sequences is an emerging research area with potentially revolutionary impact. The cost and latency of wet-lab experiments requires methods that find good sequences in few experimental rounds of large batches of sequences--a setting that off-the-shelf black-box optimization methods are ill-equipped to handle. We find that the performance of existing methods varies drastically across optimization tasks, posing a significant obstacle to real-world applications. To improve robustness, we propose Population-Based Black-Box Optimization (P3BO), which generates batches of sequences by sampling from an ensemble of methods. The number of sequences sampled from any method is proportional to the quality of sequences it previously proposed, allowing P3BO to combine the strengths of individual methods while hedging against their innate brittleness. Adapting the hyper-parameters of each of the methods online using evolutionary optimization further improves performance. Through extensive experiments on in-silico optimization tasks, we show that P3BO outperforms any single method in its population, proposing higher quality sequences as well as more diverse batches. As such, P3BO and Adaptive-P3BO are a crucial step towards deploying ML to real-world sequence design.

James Atwood, Hansa Srinivasan, Yoni Halpern et al.

Simulations of infectious disease spread have long been used to understand how epidemics evolve and how to effectively treat them. However, comparatively little attention has been paid to understanding the fairness implications of different treatment strategies -- that is, how might such strategies distribute the expected disease burden differentially across various subgroups or communities in the population? In this work, we define the precision disease control problem -- the problem of optimally allocating vaccines in a social network in a step-by-step fashion -- and we use the ML Fairness Gym to simulate epidemic control and study it from both an efficiency and fairness perspective. We then present an exploratory analysis of several different environments and discuss the fairness implications of different treatment strategies.

Yaniv Ovadia, Emily Fertig, Jie Ren et al.

Modern machine learning methods including deep learning have achieved great success in predictive accuracy for supervised learning tasks, but may still fall short in giving useful estimates of their predictive {\em uncertainty}. Quantifying uncertainty is especially critical in real-world settings, which often involve input distributions that are shifted from the training distribution due to a variety of factors including sample bias and non-stationarity. In such settings, well calibrated uncertainty estimates convey information about when a model's output should (or should not) be trusted. Many probabilistic deep learning methods, including Bayesian-and non-Bayesian methods, have been proposed in the literature for quantifying predictive uncertainty, but to our knowledge there has not previously been a rigorous large-scale empirical comparison of these methods under dataset shift. We present a large-scale benchmark of existing state-of-the-art methods on classification problems and investigate the effect of dataset shift on accuracy and calibration. We find that traditional post-hoc calibration does indeed fall short, as do several other previous methods. However, some methods that marginalize over models give surprisingly strong results across a broad spectrum of tasks.

Daniel Smilkov, Nikhil Thorat, Yannick Assogba et al.

TensorFlow.js is a library for building and executing machine learning algorithms in JavaScript. TensorFlow.js models run in a web browser and in the Node.js environment. The library is part of the TensorFlow ecosystem, providing a set of APIs that are compatible with those in Python, allowing models to be ported between the Python and JavaScript ecosystems. TensorFlow.js has empowered a new set of developers from the extensive JavaScript community to build and deploy machine learning models and enabled new classes of on-device computation. This paper describes the design, API, and implementation of TensorFlow.js, and highlights some of the impactful use cases.

D Sculley, Jasper Snoek, Alex Wiltschko

Peer review is the foundation of scientific publication, and the task of reviewing has long been seen as a cornerstone of professional service. However, the massive growth in the field of machine learning has put this community benefit under stress, threatening both the sustainability of an effective review process and the overall progress of the field. In this position paper, we argue that a tragedy of the commons outcome may be avoided by emphasizing the professional aspects of this service. In particular, we propose a rubric to hold reviewers to an objective standard for review quality. In turn, we also propose that reviewers be given appropriate incentive. As one possible such incentive, we explore the idea of financial compensation on a per-review basis. We suggest reasonable funding models and thoughts on long term effects.

Alexey A. Gritsenko, Alex D'Amour, James Atwood et al.

We introduce the BriarPatch, a pixel-space intervention that obscures sensitive attributes from representations encoded in pre-trained classifiers. The patches encourage internal model representations not to encode sensitive information, which has the effect of pushing downstream predictors towards exhibiting demographic parity with respect to the sensitive information. The net result is that these BriarPatches provide an intervention mechanism available at user level, and complements prior research on fair representations that were previously only applicable by model developers and ML experts.

Jennifer N. Wei, David Belanger, Ryan P. Adams et al.

When confronted with a substance of unknown identity, researchers often perform mass spectrometry on the sample and compare the observed spectrum to a library of previously-collected spectra to identify the molecule. While popular, this approach will fail to identify molecules that are not in the existing library. In response, we propose to improve the library's coverage by augmenting it with synthetic spectra that are predicted using machine learning. We contribute a lightweight neural network model that quickly predicts mass spectra for small molecules. Achieving high accuracy predictions requires a novel neural network architecture that is designed to capture typical fragmentation patterns from electron ionization. We analyze the effects of our modeling innovations on library matching performance and compare our models to prior machine learning-based work on spectrum prediction.

Heng-Tze Cheng, Zakaria Haque, Lichan Hong et al.

We present a framework for specifying, training, evaluating, and deploying machine learning models. Our focus is on simplifying cutting edge machine learning for practitioners in order to bring such technologies into production. Recognizing the fast evolution of the field of deep learning, we make no attempt to capture the design space of all possible model architectures in a domain- specific language (DSL) or similar configuration language. We allow users to write code to define their models, but provide abstractions that guide develop- ers to write models in ways conducive to productionization. We also provide a unifying Estimator interface, making it possible to write downstream infrastructure (e.g. distributed training, hyperparameter tuning) independent of the model implementation. We balance the competing demands for flexibility and simplicity by offering APIs at different levels of abstraction, making common model architectures available out of the box, while providing a library of utilities designed to speed up experimentation with model architectures. To make out of the box models flexible and usable across a wide range of problems, these canned Estimators are parameterized not only over traditional hyperparameters, but also using feature columns, a declarative specification describing how to interpret input data. We discuss our experience in using this framework in re- search and production environments, and show the impact on code health, maintainability, and development speed.

Brian Patton, Yannis Agiomyrgiannakis, Michael Terry et al.

Developers of text-to-speech synthesizers (TTS) often make use of human raters to assess the quality of synthesized speech. We demonstrate that we can model human raters' mean opinion scores (MOS) of synthesized speech using a deep recurrent neural network whose inputs consist solely of a raw waveform. Our best models provide utterance-level estimates of MOS only moderately inferior to sampled human ratings, as shown by Pearson and Spearman correlations. When multiple utterances are scored and averaged, a scenario common in synthesizer quality assessment, AutoMOS achieves correlations approaching those of human raters. The AutoMOS model has a number of applications, such as the ability to explore the parameter space of a speech synthesizer without requiring a human-in-the-loop.

Daniel Golovin, D. Sculley, H. Brendan McMahan et al.

We reduce the memory footprint of popular large-scale online learning methods by projecting our weight vector onto a coarse discrete set using randomized rounding. Compared to standard 32-bit float encodings, this reduces RAM usage by more than 50% during training and by up to 95% when making predictions from a fixed model, with almost no loss in accuracy. We also show that randomized counting can be used to implement per-coordinate learning rates, improving model quality with little additional RAM. We prove these memory-saving methods achieve regret guarantees similar to their exact variants. Empirical evaluation confirms excellent performance, dominating standard approaches across memory versus accuracy tradeoffs.