Google Vizier

Golovin, Daniel; Solnik, Benjamin; Moitra, Subhodeep; Kochanski, Greg; Karro, John; Sculley, D.

doi:10.1145/3097983.3098043

Cited by 380 publications

(105 citation statements)

References 12 publications

Supporting

Mentioning

101

Contrasting

Unclassified

Order By: Relevance

“…2) using Google Hypertune (Golovin et al, 2017). Hypertune is an automatic function optimizer that tries to find a minimum of a function in a bounded space.…”

Section: Resultsmentioning

confidence: 99%

“…In designing our network, we used such a system: an early version of Google Hypertune (Golovin et al, 2017). Hypertune has two components: a learning component models the effect on network performance of various hyperparameters, and an optimization component suggests new hyperparameter settings to evaluate in an attempt to find the best setting.…”

Section: Star Methodsmentioning

confidence: 99%

See 1 more Smart Citation

In Silico Labeling: Predicting Fluorescent Labels in Unlabeled Images

Christiansen

Yang

Ando

et al. 2018

Cell

540

465

View full text Add to dashboard Cite

Microscopy is a central method in life sciences. Many popular methods, such as antibody labeling, are used to add physical fluorescent labels to specific cellular constituents. However, these approaches have significant drawbacks, including inconsistency; limitations in the number of simultaneous labels because of spectral overlap; and necessary perturbations of the experiment, such as fixing the cells, to generate the measurement. Here, we show that a computational machine-learning approach, which we call "in silico labeling" (ISL), reliably predicts some fluorescent labels from transmitted-light images of unlabeled fixed or live biological samples. ISL predicts a range of labels, such as those for nuclei, cell type (e.g., neural), and cell state (e.g., cell death). Because prediction happens in silico, the method is consistent, is not limited by spectral overlap, and does not disturb the experiment. ISL generates biological measurements that would otherwise be problematic or impossible to acquire.

show abstract

“…2) using Google Hypertune (Golovin et al, 2017). Hypertune is an automatic function optimizer that tries to find a minimum of a function in a bounded space.…”

Section: Resultsmentioning

confidence: 99%

Section: Star Methodsmentioning

confidence: 99%

In Silico Labeling: Predicting Fluorescent Labels in Unlabeled Images

Christiansen

Yang

Ando

et al. 2018

Cell

540

465

View full text Add to dashboard Cite

show abstract

“…We used the Google-Vizier system for black-box optimization (Golovin et al 2017) to automatically tune our hyperparameters, including those for the input representations (e.g., number of bins, bin width), model architecture (e.g., number of fully connected layers, number of convolutional layers, kernel size), and training (e.g., dropout probability). Each Vizier "study" trained several thousand models to find the hyperparameter configuration that maximized the area under the receiver operating characteristic curve (see Section 5.1) over the validation set.…”

Section: Implementation and Training Proceduresmentioning

confidence: 99%

Identifying Exoplanets with Deep Learning: A Five-planet Resonant Chain around Kepler-80 and an Eighth Planet around Kepler-90

2018

View full text Add to dashboard Cite

NASA's Kepler Space Telescope was designed to determine the frequency of Earth-sized planets orbiting Sun-like stars, but these planets are on the very edge of the mission's detection sensitivity. Accurately determining the occurrence rate of these planets will require automatically and accurately assessing the likelihood that individual candidates are indeed planets, even at low signal-to-noise ratios. We present a method for classifying potential planet signals using deep learning, a class of machine learning algorithms that have recently become state-of-theart in a wide variety of tasks. We train a deep convolutional neural network to predict whether a given signal is a transiting exoplanet or a false positive caused by astrophysical or instrumental phenomena. Our model is highly effective at ranking individual candidates by the likelihood that they are indeed planets: 98.8% of the time it ranks plausible planet signals higher than false-positive signals in our test set. We apply our model to a new set of candidate signals that we identified in a search of known Kepler multi-planet systems. We statistically validate two new planets that are identified with high confidence by our model. One of these planets is part of a five-planet resonant chain around Kepler-80, with an orbital period closely matching the prediction by three-body Laplace relations. The other planet orbits Kepler-90, a star that was previously known to host seven transiting planets. Our discovery of an eighth planet brings Kepler-90 into a tie with our Sun as the star known to host the most planets.

show abstract

“…To quantify model uncertainty for our mortality prediction task, we explore the use of deep RNN ensembles and various Bayesian RNNs. For the deep ensembles approach, we optimize for the ideal hyperparameter values for our RNN model via black-box Bayesian optimization [19], and then train M replicas of the best model. Only the random seed differs between the replicas.…”

Section: Predictive Uncertainty Distributionsmentioning

confidence: 99%

Analyzing the role of model uncertainty for electronic health records

Dusenberry

Tran

Choi

et al. 2020

Proceedings of the ACM Conference on Health, Inference, and Learning

View full text Add to dashboard Cite

In medicine, both ethical and monetary costs of incorrect predictions can be significant, and the complexity of the problems often necessitates increasingly complex models. Recent work has shown that changing just the random seed is enough for otherwise well-tuned deep neural networks to vary in their individual predicted probabilities. In light of this, we investigate the role of model uncertainty methods in the medical domain. Using recurrent neural network (RNN) ensembles and various Bayesian RNNs, we show that population-level metrics, such as AUC-PR, AUC-ROC, log-likelihood, and calibration error, do not capture model uncertainty. Meanwhile, the presence of significant variability in patient-specific predictions and optimal decisions motivates the need for capturing model uncertainty. Understanding the uncertainty for individual patients is an area with clear clinical impact, such as determining when a model decision is likely to be brittle. We further show that RNNs with only Bayesian embeddings can be a more efficient way to capture model uncertainty compared to ensembles, and we analyze how model uncertainty is impacted across individual input features and patient subgroups. * Work done as a Google AI Resident.

show abstract

Google Vizier

Cited by 380 publications

References 12 publications

In Silico Labeling: Predicting Fluorescent Labels in Unlabeled Images

In Silico Labeling: Predicting Fluorescent Labels in Unlabeled Images

Identifying Exoplanets with Deep Learning: A Five-planet Resonant Chain around Kepler-80 and an Eighth Planet around Kepler-90

Analyzing the role of model uncertainty for electronic health records

Contact Info

Product

Resources

About