Scalable Hyperparameter Optimization with Lazy Gaussian Processes
Fraunhofer ITWM（弗劳恩霍夫工业数学研究所）；TU Kaiserslautern（凯泽斯劳滕工业大学）；Fraunhofer Center Machine Learning（弗劳恩霍夫机器学习中心）；Offenburg University（奥芬堡大学）
Raju Ram, Sabine Müller, Franz-Josef Pfreundt, Nicolas R. Gauger, Janis Keuper
备注：
14 pages; 6 figures; 4 tables; Accepted in proceedings of MLHPC 2019: Fifth International Workshop on Machine Learning in High Performance Computing Environments, Super Computing Conference 2019, Denver, Colorado
Most machine learning methods require careful selection of hyper-parameters in order to train a high performing model with good generalization abilities. Hence, several automatic selection algorithms have been introduced to overcome tedious manual (try and error) tuning of these parameters. Due to its very high sample efficiency, Bayesian Optimization over a Gaussian Processes modeling of the parameter space has become the method of choice. Unfortunately, this approach suffers from a cubic compute complexity due to underlying Cholesky factorization, which makes it very hard to be scaled beyond a small number of sampling steps. In this paper, we present a novel, highly accurate approximation of the underlying Gaussian Process. Reducing its computational complexity from cubic to quadratic allows an efficient strong scaling of Bayesian Optimization while outperforming the previous approach regarding optimization accuracy. The first experiments show speedups of a factor of 162 in single node and further speed up by a factor of 5 in a parallel environment.
University of Colorado Boulder（科罗拉多大学博尔德分校）；Lawrence Livermore National Laboratory（劳伦斯利弗莫尔国家实验室）
Gaussian processes (GPs) are Bayesian non-parametric models popular in a variety of applications due to their accuracy and native uncertainty quantification (UQ). Tuning GP hyperparameters is critical to ensure the validity of prediction accuracy and uncertainty; uniquely estimating multiple hyperparameters in, e.g. the Matern kernel can also be a significant challenge. Moreover, training GPs on large-scale datasets is a highly active area of research: traditional maximum likelihood hyperparameter training requires quadratic memory to form the covariance matrix and has cubic training complexit…
In recent years, leveraging parallel and distributed computational resources has become essential to solve problems of high computational cost. Bayesian optimization (BO) has shown attractive results in those expensive-to-evaluate problems such as hyperparameter optimization of machine learning algorithms. While many parallel BO methods have been developed to search efficiently utilizing these computational resources, these methods assumed synchronous settings or were not scalable. In this paper, we propose a simple and scalable BO method for asynchronous parallel settings. Experiments are car…
PriorCVAE: scalable MCMC parameter inference with Bayesian deep generative modelling
University of Oxford（牛津大学）；Aalto University（阿尔托大学）；Institute of Science and Technology Austria（奥地利科学技术研究所）；University of Copenhagen（哥本哈根大学）；Imperial College London（伦敦帝国学院）
Gaussian processes (GPs) are non-linear probabilistic models popular in many applications. However, naïve GP realizations require quadratic memory to store the covariance matrix and cubic computation to perform inference or evaluate the likelihood function. These bottlenecks have driven much investment in the development of approximate GP alternatives that scale to the large data sizes common in modern data-driven applications. We present in this manuscript MuyGPs, a novel efficient GP hyperparameter estimation method. MuyGPs builds upon prior methods that take advantage of the nearest neighbo…
Bayesian Optimisation (BO) refers to a suite of techniques for global optimisation of expensive black box functions, which use introspective Bayesian models of the function to efficiently search for the optimum. While BO has been applied successfully in many applications, modern optimisation tasks usher in new challenges where conventional methods fail spectacularly. In this work, we present Dragonfly, an open source Python library for scalable and robust BO. Dragonfly incorporates multiple recently developed methods that allow BO to be applied in challenging real world settings; these include…
University of Tübingen（图宾根大学）；Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所）；Columbia University（哥伦比亚大学）；University of Pennsylvania（宾夕法尼亚大学）
Gaussian process hyperparameter optimization requires linear solves with, and log-determinants of, large kernel matrices. Iterative numerical techniques are becoming popular to scale to larger datasets, relying on the conjugate gradient method (CG) for the linear solves and stochastic trace estimation for the log-determinant. This work introduces new algorithmic and theoretical insights for preconditioning these computations. While preconditioning is well understood in the context of CG, we demonstrate that it can also accelerate convergence and reduce variance of the estimates for the log-det…
Tuning complex machine learning systems is challenging. Machine learning typically requires to set hyperparameters, be it regularization, architecture, or optimization parameters, whose tuning is critical to achieve good predictive performance. To democratize access to machine learning systems, it is essential to automate the tuning. This paper presents Amazon SageMaker Automatic Model Tuning (AMT), a fully managed system for gradient-free optimization at scale. AMT finds the best version of a trained machine learning model by repeatedly evaluating it with different hyperparameter configuratio…
Numerous engineering problems of interest to the industry are often characterized by expensive black-box objective experiments or computer simulations. Obtaining insight into the problem or performing subsequent optimizations requires hundreds of thousands of evaluations of the objective function which is most often a practically unachievable task. Gaussian Process (GP) surrogate modeling replaces the expensive function with a cheap-to-evaluate data-driven probabilistic model. While the GP does not assume a functional form of the problem, it is defined by a set of parameters, called hyperparam…
Scalable Gaussian Process methods are computationally attractive, yet introduce modeling biases that require rigorous study. This paper analyzes two common techniques: early truncated conjugate gradients (CG) and random Fourier features (RFF). We find that both methods introduce a systematic bias on the learned hyperparameters: CG tends to underfit while RFF tends to overfit. We address these issues using randomized truncation estimators that eliminate bias in exchange for increased variance. In the case of RFF, we show that the bias-to-variance conversion is indeed a trade-off: the additional…
Cornell University（康奈尔大学）；University of Delaware（特拉华大学）；SigOpt
The Bayesian transformed Gaussian process (BTG) model, proposed by Kedem and Oliviera, is a fully Bayesian counterpart to the warped Gaussian process (WGP) and marginalizes out a joint prior over input warping and kernel hyperparameters. This fully Bayesian treatment of hyperparameters often provides more accurate regression estimates and superior uncertainty propagation, but is prohibitively expensive. The BTG posterior predictive distribution, itself estimated through high-dimensional integration, must be inverted in order to perform model prediction. To make the Bayesian approach practical …
We introduce a kernel approximation strategy that enables computation of the Gaussian process log marginal likelihood and all hyperparameter derivatives in $\mathcal{O}(p)$ time. Our GRIEF kernel consists of $p$ eigenfunctions found using a Nystrom approximation from a dense Cartesian product grid of inducing points. By exploiting algebraic properties of Kronecker and Khatri-Rao tensor products, computational complexity of the training procedure can be practically independent of the number of inducing points. This allows us to use arbitrarily many inducing points to achieve a globally accurate…
National University of Singapore（新加坡国立大学）；Shenzhen University（深圳大学）
This paper presents a variational Bayesian kernel selection (VBKS) algorithm for sparse Gaussian process regression (SGPR) models. In contrast to existing GP kernel selection algorithms that aim to select only one kernel with the highest model evidence, our proposed VBKS algorithm considers the kernel as a random variable and learns its belief from data such that the uncertainty of the kernel can be interpreted and exploited to avoid overconfident GP predictions. To achieve this, we represent the probabilistic kernel as an additional variational variable in a variational inference (VI) framewo…
Variable selection in Gaussian processes (GPs) is typically undertaken by thresholding the inverse lengthscales of automatic relevance determination kernels, but in high-dimensional datasets this approach can be unreliable. A more probabilistically principled alternative is to use spike and slab priors and infer a posterior probability of variable inclusion. However, existing implementations in GPs are very costly to run in both high-dimensional and large-$n$ datasets, or are only suitable for unsupervised settings with specific kernels. As such, we develop a fast and scalable variational infe…
Deep neural networks (NNs) are powerful black box predictors that have recently achieved impressive performance on a wide spectrum of tasks. Quantifying predictive uncertainty in NNs is a challenging and yet unsolved problem. Bayesian NNs, which learn a distribution over weights, are currently the state-of-the-art for estimating predictive uncertainty; however these require significant modifications to the training procedure and are computationally expensive compared to standard (non-Bayesian) NNs. We propose an alternative to Bayesian NNs that is simple to implement, readily parallelizable, r…
As a non-parametric Bayesian model which produces informative predictive distribution, Gaussian process (GP) has been widely used in various fields, like regression, classification and optimization. The cubic complexity of standard GP however leads to poor scalability, which poses challenges in the era of big data. Hence, various scalable GPs have been developed in the literature in order to improve the scalability while retaining desirable prediction accuracy. This paper devotes to investigating the methodological characteristics and performance of representative global and local scalable GPs…
University of Cambridge（剑桥大学）；Alan Turing Institute（艾伦·图灵研究所）
Machine learning training methods depend plentifully and intricately on hyperparameters, motivating automated strategies for their optimisation. Many existing algorithms restart training for each new hyperparameter choice, at considerable computational cost. Some hypergradient-based one-pass methods exist, but these either cannot be applied to arbitrary optimiser hyperparameters (such as learning rates and momenta) or take several times longer to train than their base models. We extend these existing methods to develop an approximate hypergradient-based hyperparameter optimiser which is applic…
Scalable Logistic Gaussian Process Density Regression with Kinetic Langevin Sampling
Nanyang Technological University（南洋理工大学）；Eötvös Loránd University (ELTE)（厄特沃什·罗兰大学（ELTE））；HUN-REN Institute for Computer Science and Control (SZTAKI)（匈牙利国家研究网络计算机科学与控制研究所（SZTAKI））
Large multilayer neural networks trained with backpropagation have recently achieved state-of-the-art results in a wide range of problems. However, using backprop for neural net learning still has some disadvantages, e.g., having to tune a large number of hyperparameters to the data, lack of calibrated probabilistic predictions, and a tendency to overfit the training data. In principle, the Bayesian approach to learning neural networks does not have these problems. However, existing Bayesian techniques lack scalability to large dataset and network sizes. In this work we present a novel scalabl…
While recent work on conjugate gradient methods and Lanczos decompositions have achieved scalable Gaussian process inference with highly accurate point predictions, in several implementations these iterative methods appear to struggle with numerical instabilities in learning kernel hyperparameters, and poor test likelihoods. By investigating CG tolerance, preconditioner rank, and Lanczos decomposition rank, we provide a particularly simple prescription to correct these issues: we recommend that one should use a small CG tolerance ($ε\leq 0.01$) and a large root decomposition size ($r \geq 5000…
Bayesian optimization is an effective methodology for the global optimization of functions with expensive evaluations. It relies on querying a distribution over functions defined by a relatively cheap surrogate model. An accurate model for this distribution over functions is critical to the effectiveness of the approach, and is typically fit using Gaussian processes (GPs). However, since GPs scale cubically with the number of observations, it has been challenging to handle objectives whose optimization requires many evaluations, and as such, massively parallelizing the optimization. In this wo…
Aalto University（阿尔托大学）；IDSIA, Dalle Molle Institute for Artificial Intelligence, USI-SUPSI（IDSIA，达勒·莫勒人工智能研究所（USI-SUPSI））
Hyperparameter selection generally relies on running multiple full training trials, with selection based on validation set performance. We propose a gradient-based approach for locally adjusting hyperparameters during training of the model. Hyperparameters are adjusted so as to make the model parameter gradients, and hence updates, more advantageous for the validation cost. We explore the approach for tuning regularization hyperparameters and find that in experiments on MNIST, SVHN and CIFAR-10, the resulting regularization levels are within the optimal regions. The additional computational co…
Lawrence Livermore National Laboratory（劳伦斯利弗莫尔国家实验室）
Gaussian processes (GPs) are Bayesian non-parametric models useful in a myriad of applications. Despite their popularity, the cost of GP predictions (quadratic storage and cubic complexity with respect to the number of training points) remains a hurdle in applying GPs to large data. We present a fast posterior mean prediction algorithm called FastMuyGPs to address this shortcoming. FastMuyGPs is based upon the MuyGPs hyperparameter estimation algorithm and utilizes a combination of leave-one-out cross-validation, batching, nearest neighbors sparsification, and precomputation to provide scalabl…
Deep generative models have achieved great success in unsupervised learning with the ability to capture complex nonlinear relationships between latent generating factors and observations. Among them, a factorized hierarchical variational autoencoder (FHVAE) is a variational inference-based model that formulates a hierarchical generative process for sequential data. Specifically, an FHVAE model can learn disentangled and interpretable representations, which have been proven useful for numerous speech applications, such as speaker verification, robust speech recognition, and voice conversion. Ho…
Causal Local States: Scalable Simultaneous Causal Network Inference and Forecasting for Dynamical Systems
Ludwig-Maximilians-Universität München（慕尼黑大学）；Institut für Theoretische Physik, Universität zu Köln（科隆大学理论物理研究所）；Institut für KI-Sicherheit, Deutsches Zentrum für Luft- und Raumfahrt (DLR)（德国航空航天中心人工智能安全研究所）；Entrox Systems（恩特罗克斯系统公司）；Institut für Frontier Materials auf der Erde und im Weltraum, Deutsches Zentrum für Luft- und Raumfahrt (DLR)（德国航空航天中心地球与太空前沿材料研究所）
Computer vision is experiencing an AI renaissance, in which machine learning models are expediting important breakthroughs in academic research and commercial applications. Effectively training these models, however, is not trivial due in part to hyperparameters: user-configured values that control a model's ability to learn from data. Existing hyperparameter optimization methods are highly parallel but make no effort to balance the search across heterogeneous hardware or to prioritize searching high-impact spaces. In this paper, we introduce a framework for massively Scalable Hardware-Aware D…
Oak Ridge National Laboratory（橡树岭国家实验室）；Los Alamos National Laboratory（洛斯阿拉莫斯国家实验室）
We propose a new scalable method to optimize the architecture of an artificial neural network. The proposed algorithm, called Greedy Search for Neural Network Architecture, aims to determine a neural network with minimal number of layers that is at least as performant as neural networks of the same structure identified by other hyperparameter search algorithms in terms of accuracy and computational cost. Numerical results performed on benchmark datasets show that, for these datasets, our method outperforms state-of-the-art hyperparameter optimization algorithms in terms of attainable predictiv…
Ant Financial Services Group（蚂蚁金融服务集团）；Cisco Systems（思科系统公司）；Caicloud（才云科技）；Google（谷歌）；IBM（国际商业机器公司）
In this paper, we introduce Katib: a scalable, cloud-native, and production-ready hyperparameter tuning system that is agnostic of the underlying machine learning framework. Though there are multiple hyperparameter tuning systems available, this is the first one that caters to the needs of both users and administrators of the system. We present the motivation and design of the system and contrast it with existing hyperparameter tuning systems, especially in terms of multi-tenancy, scalability, fault-tolerance, and extensibility. It can be deployed on local machines, or hosted as a service in o…
University of Freiburg（弗莱堡大学）；Bosch Center for Artificial Intelligence（博世人工智能中心）
Modern machine learning algorithms crucially rely on several design decisions to achieve strong performance, making the problem of Hyperparameter Optimization (HPO) more important than ever. Here, we combine the advantages of the popular bandit-based HPO method Hyperband (HB) and the evolutionary search approach of Differential Evolution (DE) to yield a new HPO method which we call DEHB. Comprehensive results on a very broad range of HPO problems, as well as a wide range of tabular benchmarks from neural architecture search, demonstrate that DEHB achieves strong performance far more robustly t…
ETH Zurich（苏黎世联邦理工学院）；Max Planck ETH Center for Learning Systems（马克斯·普朗克-苏黎世联邦理工学习系统中心）；Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所）；University of Cambridge（剑桥大学）；RIKEN Center for Advanced Intelligence Project（理化学研究所先进智能项目中心）
Marginal-likelihood based model-selection, even though promising, is rarely used in deep learning due to estimation difficulties. Instead, most approaches rely on validation data, which may not be readily available. In this work, we present a scalable marginal-likelihood estimation method to select both hyperparameters and network architectures, based on the training data alone. Some hyperparameters can be estimated online during training, simplifying the procedure. Our marginal-likelihood estimate is based on Laplace's method and Gauss-Newton approximations to the Hessian, and it outperforms …
Temporal variations of apparent magnitude, called light curves, are observational statistics of interest captured by telescopes over long periods of time. Light curves afford the exploration of Space Domain Awareness (SDA) objectives such as object identification or pose estimation as latent variable inference problems. Ground-based observations from commercial off the shelf (COTS) cameras remain inexpensive compared to higher precision instruments, however, limited sensor availability combined with noisier observations can produce gappy time-series data that can be difficult to model. These e…
University of Toronto（多伦多大学）；Vector Institute（向量研究所）
Hyperparameter optimization can be formulated as a bilevel optimization problem, where the optimal parameters on the training set depend on the hyperparameters. We aim to adapt regularization hyperparameters for neural networks by fitting compact approximations to the best-response function, which maps hyperparameters to optimal weights and biases. We show how to construct scalable best-response approximations for neural networks by modeling the best-response as a single network whose hidden units are gated conditionally on the regularizer. We justify this approximation by showing the exact be…
Tufts University（塔夫茨大学）；University of Sao Paulo（圣保罗大学）；University of Utrecht（乌得勒支大学）；Indiana University Bloomington（印第安纳大学伯明顿分校）；Novartis（诺华公司）
Projection algorithms such as t-SNE or UMAP are useful for the visualization of high dimensional data, but depend on hyperparameters which must be tuned carefully. Unfortunately, iteratively recomputing projections to find the optimal hyperparameter value is computationally intensive and unintuitive due to the stochastic nature of these methods. In this paper we propose HyperNP, a scalable method that allows for real-time interactive hyperparameter exploration of projection methods by training neural network approximations. HyperNP can be trained on a fraction of the total data instances and h…
Streamlining Ocean Dynamics Modeling with Fourier Neural Operators: A Multiobjective Hyperparameter and Architecture Optimization Approach
Argonne National Laboratory（阿贡国家实验室）；Florida State University（佛罗里达州立大学）；Université Paris-Saclay（巴黎萨克雷大学）；Los Alamos National Laboratory（洛斯阿拉莫斯国家实验室）；Oak Ridge National Laboratory（橡树岭国家实验室）
Outlier detection (OD) literature exhibits numerous algorithms as it applies to diverse domains. However, given a new detection task, it is unclear how to choose an algorithm to use, nor how to set its hyperparameter(s) (HPs) in unsupervised settings. HP tuning is an ever-growing problem with the arrival of many new detectors based on deep learning, which usually come with a long list of HPs. Surprisingly, the issue of model selection in the outlier mining literature has been "the elephant in the room"; a significant factor in unlocking the utmost potential of deep methods, yet little said or …
The purpose of this study is to introduce new design-criteria for next-generation hyperparameter optimization software. The criteria we propose include (1) define-by-run API that allows users to construct the parameter search space dynamically, (2) efficient implementation of both searching and pruning strategies, and (3) easy-to-setup, versatile architecture that can be deployed for various purposes, ranging from scalable distributed computing to light-weight experiment conducted via interactive interface. In order to prove our point, we will introduce Optuna, an optimization software which i…
Argonne National Laboratory（阿贡国家实验室）；Lawrence Berkeley National Laboratory（劳伦斯伯克利国家实验室）
Deep-learning-based data-driven forecasting methods have produced impressive results for traffic forecasting. A major limitation of these methods, however, is that they provide forecasts without estimates of uncertainty, which are critical for real-time deployments. We focus on a diffusion convolutional recurrent neural network (DCRNN), a state-of-the-art method for short-term traffic forecasting. We develop a scalable deep ensemble approach to quantify uncertainties for DCRNN. Our approach uses a scalable Bayesian optimization method to perform hyperparameter optimization, selects a set of hi…
KAIST（韩国科学技术院）；AITRICS（艾特里克斯）；Lunit（鲁尼特）；University of Edinburgh（爱丁堡大学）；Samsung AI Centre Cambridge（三星剑桥人工智能中心）
Many gradient-based meta-learning methods assume a set of parameters that do not participate in inner-optimization, which can be considered as hyperparameters. Although such hyperparameters can be optimized using the existing gradient-based hyperparameter optimization (HO) methods, they suffer from the following issues. Unrolled differentiation methods do not scale well to high-dimensional hyperparameters or horizon length, Implicit Function Theorem (IFT) based methods are restrictive for online optimization, and short horizon approximations suffer from short horizon bias. In this work, we pro…
Technion - Israel Institute of Technology（以色列理工学院）；Habana-Labs（哈瓦那实验室）
Background: Recent developments have made it possible to accelerate neural networks training significantly using large batch sizes and data parallelism. Training in an asynchronous fashion, where delay occurs, can make training even more scalable. However, asynchronous training has its pitfalls, mainly a degradation in generalization, even after convergence of the algorithm. This gap remains not well understood, as theoretical analysis so far mainly focused on the convergence rate of asynchronous methods. Contributions: We examine asynchronous training from the perspective of dynamical stabili…
University of Toronto（多伦多大学）；RBC Capital Markets（加拿大皇家银行资本市场）
Advances in Reinforcement Learning (RL) have demonstrated data efficiency and optimal control over large state spaces at the cost of scalable performance. Genetic methods, on the other hand, provide scalability but depict hyperparameter sensitivity towards evolutionary operations. However, a combination of the two methods has recently demonstrated success in scaling RL agents to high-dimensional action spaces. Parallel to recent developments, we present the Evolution-based Soft Actor-Critic (ESAC), a scalable RL algorithm. We abstract exploration from exploitation by combining Evolution Strate…
