Algorithm parameters, in particular hyperparameters of machine learning algorithms, can substantially impact their performance. To support users in determining well-performing hyperparameter configurations for their algorithms, datasets and applications at hand, SMAC3 offers a robust and flexible framework for Bayesian Optimization, which can improve performance within a few evaluations. It offers several facades and pre-sets for typical use cases, such as optimizing hyperparameters, solving low dimensional continuous (artificial) global optimization problems and configuring algorithms to perform well across multiple problem instances. The SMAC3 package is available under a permissive BSD-license at https://github.com/automl/SMAC3.
Bayesian Optimization (BO) is a method for globally optimizing black-box functions. While BO has been successfully applied to many scenarios, developing effective BO algorithms that scale to functions with high-dimensional domains is still a challenge. Optimizing such functions by vanilla BO is extremely time-consuming. Alternative strategies for high-dimensional BO that are based on the idea of embedding the high-dimensional space to the one with low dimension are sensitive to the choice of the embedding dimension, which needs to be pre-specified. We develop a new computationally efficient high-dimensional BO method that exploits variable selection. Our method is able to automatically learn axis-aligned sub-spaces, i.e. spaces containing selected variables, without the demand of any pre-specified hyperparameters. We theoretically analyze the computational complexity of our algorithm and derive the regret bound. We empirically show the efficacy of our method on several synthetic and real problems.
Bayesian optimization (BO) has become a popular strategy for global optimization of expensive real-world functions. Contrary to a common expectation that BO is suited to optimizing black-box functions, it actually requires domain knowledge about those functions to deploy BO successfully. Such domain knowledge often manifests in Gaussian process (GP) priors that specify initial beliefs on functions. However, even with expert knowledge, it is non-trivial to quantitatively define a prior. This is especially true for hyperparameter tuning problems on complex machine learning models, where landscapes of tuning objectives are often difficult to comprehend. We seek an alternative practice for setting these functional priors. In particular, we consider the scenario where we have data from similar functions that allow us to pre-train a tighter distribution a priori. We detail what pre-training entails for GPs using a KL divergence based loss function, and propose a new pre-training based BO framework named HyperBO. Theoretically, we show bounded posterior predictions and near-zero regrets for HyperBO without assuming the "ground truth" GP prior is known. To verify our approach in realistic setups, we collect a large multi-task hyperparameter tuning dataset by training tens of thousands of configurations of near-state-of-the-art deep learning models on popular image and text datasets, as well as a protein sequence dataset. Our results show that on average, HyperBO is able to locate good hyperparameters at least 3 times more efficiently than the best competing methods on both our new tuning dataset and existing multi-task BO benchmarks.
Building surrogate models is one common approach when we attempt to learn unknown black-box functions. Bayesian optimization provides a framework which allows us to build surrogate models based on sequential samples drawn from the function and find the optimum. Tuning algorithmic parameters to optimize the performance of large, complicated "black-box" application codes is a specific important application, which aims at finding the optima of black-box functions. Within the Bayesian optimization framework, the Gaussian process model produces smooth or continuous sample paths. However, the black-box function in the tuning problem is often non-smooth. This difficult tuning problem is worsened by the fact that we usually have limited sequential samples from the black-box function. Motivated by these issues encountered in tuning, we propose a novel additive Gaussian process model called clustered Gaussian process (cGP), where the additive components are induced by clustering. In the examples we studied, the performance can be improved by as much as 90% among repetitive experiments. By using this surrogate model, we want to capture the non-smoothness of the black-box function. In addition to an algorithm for constructing this model, we also apply the model to several artificial and real applications to evaluate it.
Being able to reply with a related, fluent, and informative response is an indispensable requirement for building high-quality conversational agents. In order to generate better responses, some approaches have been proposed, such as feeding extra information by collecting large-scale datasets with human annotations, designing neural conversational models (NCMs) with complex architecture and loss functions, or filtering out untrustworthy samples based on a dialogue attribute, e.g., Relatedness or Genericness. In this paper, we follow the third research branch and present a data filtering method for open-domain dialogues, which identifies untrustworthy samples from training data with a quality measure that linearly combines seven dialogue attributes. The attribute weights are obtained via Bayesian Optimization (BayesOpt) that aims to optimize an objective function for dialogue generation iteratively on the validation set. Then we score training samples with the quality measure, sort them in descending order, and filter out those at the bottom. Furthermore, to accelerate the "filter-train-evaluate" iterations involved in BayesOpt on large-scale datasets, we propose a training framework that integrates maximum likelihood estimation (MLE) and negative training method (NEG). The training method updates parameters of a trained NCMs on two small sets with newly maintained and removed samples, respectively. Specifically, MLE is applied to maximize the log-likelihood of newly maintained samples, while NEG is used to minimize the log-likelihood of newly removed ones. Experimental results on two datasets show that our method can effectively identify untrustworthy samples, and NCMs trained on the filtered datasets achieve better performance.
Modern deep learning (DL) applications are built using DL libraries and frameworks such as TensorFlow and PyTorch. These frameworks have complex parameters and tuning them to obtain good training and inference performance is challenging for typical users, such as DL developers and data scientists. Manual tuning requires deep knowledge of the user-controllable parameters of DL frameworks as well as the underlying hardware. It is a slow and tedious process, and it typically delivers sub-optimal solutions. In this paper, we treat the problem of tuning parameters of DL frameworks to improve training and inference performance as a black-box optimization problem. We then investigate applicability and effectiveness of Bayesian optimization (BO), genetic algorithm (GA), and Nelder-Mead simplex (NMS) to tune the parameters of TensorFlow's CPU backend. While prior work has already investigated the use of Nelder-Mead simplex for a similar problem, it does not provide insights into the applicability of other more popular algorithms. Towards that end, we provide a systematic comparative analysis of all three algorithms in tuning TensorFlow's CPU backend on a variety of DL models. Our findings reveal that Bayesian optimization performs the best on the majority of models. There are, however, cases where it does not deliver the best results.
The decoding of brain signals recorded via, e.g., an electroencephalogram, using machine learning is key to brain-computer interfaces (BCIs). Stimulation parameters or other experimental settings of the BCI protocol typically are chosen according to the literature. The decoding performance directly depends on the choice of parameters, as they influence the elicited brain signals and optimal parameters are subject-dependent. Thus a fast and automated selection procedure for experimental parameters could greatly improve the usability of BCIs. We evaluate a standalone random search and a combined Bayesian optimization with random search in a closed-loop auditory event-related potential protocol. We aimed at finding the individually best stimulation speed -- also known as stimulus onset asynchrony (SOA) -- that maximizes the classification performance of a regularized linear discriminant analysis. To make the Bayesian optimization feasible under noise and the time pressure posed by an online BCI experiment, we first used offline simulations to initialize and constrain the internal optimization model. Then we evaluated our approach online with 13 healthy subjects. We could show that for 8 out of 13 subjects, the proposed approach using Bayesian optimization succeeded to select the individually optimal SOA out of multiple evaluated SOA values. Our data suggests, however, that subjects were influenced to very different degrees by the SOA parameter. This makes the automatic parameter selection infeasible for subjects where the influence is limited. Our work proposes an approach to exploit the benefits of individualized experimental protocols and evaluated it in an auditory BCI. When applied to other experimental parameters our approach could enhance the usability of BCI for different target groups -- specifically if an individual disease progress may prevent the use of standard parameters.
As the complexity of control systems increases, the need for systematic methods to guarantee their efficacy grows as well. However, direct testing of these systems is oftentimes costly, difficult, or impractical. As a result, the test and evaluation ideal would be to verify the efficacy of a system simulator and use this verification result to make a statement on true system performance. This paper formalizes that performance translation for a specific class of desired system behaviors. In that vein, our contribution is twofold. First, we detail a variant on existing Bayesian Optimization Algorithms that identifies minimal upper bounds to maximization problems, with some minimum probability. Second, we use this Algorithm to $i)$ lower bound the minimum simulator robustness and $ii)$ upper bound the expected deviance between true and simulated systems. Then, for the specific class of desired behaviors studied, we leverage these bounds to lower bound the minimum true system robustness, without directly testing the true system. Finally, we compare a high-fidelity ROS simulator of a Segway, with a significantly noisier version of itself, and show that our probabilistic verification bounds are indeed satisfied.
This paper considers the open challenge of identifying complete, concise, and explainable quantitative microstructure representations for disordered heterogeneous material systems. Completeness and conciseness have been achieved through existing data-driven methods, e.g., deep generative models, which, however, do not provide mathematically explainable latent representations. This study investigates representations composed of three-point correlation functions, which are a special type of spatial convolutions. We show that a variety of microstructures can be characterized by a concise subset of three-point correlations, and the identification of such subsets can be achieved by Bayesian optimization. Lastly, we show that the proposed representation can directly be used to compute material properties based on the effective medium theory.
A large body of literature has proved that the Bayesian optimization framework is especially efficient and effective in analog circuit synthesis. However, most of the previous research works only focus on designing informative surrogate models or efficient acquisition functions. Even if searching for the global optimum over the acquisition function surface is itself a difficult task, it has been largely ignored. In this paper, we propose a fast and robust Bayesian optimization approach via one-dimensional subspaces for analog circuit synthesis. By solely focusing on optimizing one-dimension subspaces at each iteration, we greatly reduce the computational overhead of the Bayesian optimization framework while safely maximizing the acquisition function. By combining the benefits of different dimension selection strategies, we adaptively balancing between searching globally and locally. By leveraging the batch Bayesian optimization framework, we further accelerate the optimization procedure by making full use of the hardware resources. Experimental results quantitatively show that our proposed algorithm can accelerate the optimization procedure by up to 9x and 38x compared to LP-EI and REMBOpBO respectively when the batch size is 15.
Data augmentation aims to enrich training samples for alleviating the overfitting issue in low-resource or class-imbalanced situations. Traditional methods first devise task-specific operations such as Synonym Substitute, then preset the corresponding parameters such as the substitution rate artificially, which require a lot of prior knowledge and are prone to fall into the sub-optimum. Besides, the number of editing operations is limited in the previous methods, which decreases the diversity of the augmented data and thus restricts the performance gain. To overcome the above limitations, we propose a framework named Text AutoAugment (TAA) to establish a compositional and learnable paradigm for data augmentation. We regard a combination of various operations as an augmentation policy and utilize an efficient Bayesian Optimization algorithm to automatically search for the best policy, which substantially improves the generalization capability of models. Experiments on six benchmark datasets show that TAA boosts classification accuracy in low-resource and class-imbalanced regimes by an average of 8.8% and 9.7%, respectively, outperforming strong baselines.
There is a wave of interest in using unsupervised neural networks for solving differential equations. The existing methods are based on feed-forward networks, {while} recurrent neural network differential equation solvers have not yet been reported. We introduce an unsupervised reservoir computing (RC), an echo-state recurrent neural network capable of discovering approximate solutions that satisfy ordinary differential equations (ODEs). We suggest an approach to calculate time derivatives of recurrent neural network outputs without using backpropagation. The internal weights of an RC are fixed, while only a linear output layer is trained, yielding efficient training. However, RC performance strongly depends on finding the optimal hyper-parameters, which is a computationally expensive process. We use Bayesian optimization to efficiently discover optimal sets in a high-dimensional hyper-parameter space and numerically show that one set is robust and can be used to solve an ODE for different initial conditions and time ranges. A closed-form formula for the optimal output weights is derived to solve first order linear equations in a backpropagation-free learning process. We extend the RC approach by solving nonlinear system of ODEs using a hybrid optimization method consisting of gradient descent and Bayesian optimization. Evaluation of linear and nonlinear systems of equations demonstrates the efficiency of the RC ODE solver.
In this paper, we propose an approach to address the problems with ambiguity in tuning the process and observation noises for a discrete-time linear Kalman filter. Conventional approaches to tuning (e.g. using normalized estimation error squared and covariance minimization) compute empirical measures of filter performance and the parameter are selected manually or selected using some kind of optimization algorithm to maximize these measures of performance. However, there are two challenges with this approach. First, in theory, many of these measures do not guarantee a unique solution due to observability issues. Second, in practice, empirically computed statistical quantities can be very noisy due to a finite number of samples. We propose a method to overcome these limitations. Our method has two main parts to it. The first is to ensure that the tuning problem has a single unique solution. We achieve this by simultaneously tuning the filter over multiple different prediction intervals. Although this yields a unique solution, practical issues (such as sampling noise) mean that it cannot be directly applied. Therefore, we use Bayesian Optimization. This technique handles noisy data and the local minima that it introduces.
Many applications in important problem domains such as machine learning and computer vision are streaming applications that take a sequence of inputs over time. It is challenging to find knob settings that optimize the run-time performance of such applications because the optimal knob settings are usually functions of inputs, computing platforms, time as well as user's requirements, which can be very diverse. Most prior works address this problem by offline profiling followed by training models for control. However, profiling-based approaches incur large overhead before execution; it is also difficult to redeploy them in other run-time configurations. In this paper, we propose Sonic, a sampling-based online controller for long-running streaming applications that does not require profiling ahead of time. Within each phase of a streaming application's execution, Sonic utilizes the beginning portion to sample the knob space strategically and aims to pick the optimal knob setting for the rest of the phase, given a user-specified constrained optimization problem. A hybrid approach of machine learning regressions and Bayesian optimization are used for better overall sampling choices. Sonic is implemented independent of application, device, input, performance objective and constraints. We evaluate Sonic on traditional parallel benchmarks as well as on deep learning inference benchmarks across multiple platforms. Our experiments show that when using Sonic to control knob settings, application run-time performance is only 5.3% less than if optimal knob settings were used, demonstrating that Sonic is able to find near-optimal knob settings under diverse run-time configurations without prior knowledge quickly.
In this paper we aim to provide analysis and insights (often based on visualization), which explain the beneficial effects of on-line decision making on top of off-line training. In particular, through a unifying abstract mathematical framework, we show that the principal AlphaZero/TD-Gammon ideas of approximation in value space and rollout apply very broadly to deterministic and stochastic optimal control problems, involving both discrete and continuous search spaces. Moreover, these ideas can be effectively integrated with other important methodologies such as model predictive control, adaptive control, decentralized control, discrete and Bayesian optimization, neural network-based value and policy approximations, and heuristic algorithms for discrete optimization.
Glioblastoma is profoundly heterogeneous in regional microstructure and vasculature. Characterizing the spatial heterogeneity of glioblastoma could lead to more precise treatment. With unsupervised learning techniques, glioblastoma MRI-derived radiomic features have been widely utilized for tumor sub-region segmentation and survival prediction. However, the reliability of algorithm outcomes is often challenged by both ambiguous intermediate process and instability introduced by the randomness of clustering algorithms, especially for data from heterogeneous patients. In this paper, we propose an adaptive unsupervised learning approach for efficient MRI intra-tumor partitioning and glioblastoma survival prediction. A novel and problem-specific Feature-enhanced Auto-Encoder (FAE) is developed to enhance the representation of pairwise clinical modalities and therefore improve clustering stability of unsupervised learning algorithms such as K-means. Moreover, the entire process is modelled by the Bayesian optimization (BO) technique with a custom loss function that the hyper-parameters can be adaptively optimized in a reasonably few steps. The results demonstrate that the proposed approach can produce robust and clinically relevant MRI sub-regions and statistically significant survival predictions.
The world is going through a challenging phase due to the disastrous effect caused by the COVID-19 pandemic on the healthcare system and the economy. The rate of spreading, post-COVID-19 symptoms, and the occurrence of new strands of COVID-19 have put the healthcare systems in disruption across the globe. Due to this, the task of accurately screening COVID-19 cases has become of utmost priority. Since the virus infects the respiratory system, Chest X-Ray is an imaging modality that is adopted extensively for the initial screening. We have performed a comprehensive study that uses CXR images to identify COVID-19 cases and realized the necessity of having a more generalizable model. We utilize MobileNetV2 architecture as the feature extractor and integrate it into Capsule Networks to construct a fully automated and lightweight model termed as MobileCaps. MobileCaps is trained and evaluated on the publicly available dataset with the model ensembling and Bayesian optimization strategies to efficiently classify CXR images of patients with COVID-19 from non-COVID-19 pneumonia and healthy cases. The proposed model is further evaluated on two additional RT-PCR confirmed datasets to demonstrate the generalizability. We also introduce MobileCaps-S and leverage it for performing severity assessment of CXR images of COVID-19 based on the Radiographic Assessment of Lung Edema (RALE) scoring technique. Our classification model achieved an overall recall of 91.60, 94.60, 92.20, and a precision of 98.50, 88.21, 92.62 for COVID-19, non-COVID-19 pneumonia, and healthy cases, respectively. Further, the severity assessment model attained an R$^2$ coefficient of 70.51. Owing to the fact that the proposed models have fewer trainable parameters than the state-of-the-art models reported in the literature, we believe our models will go a long way in aiding healthcare systems in the battle against the pandemic.
Predicting clinical outcomes from medical images using quantitative features (``radiomics'') requires many method design choices, Currently, in new clinical applications, finding the optimal radiomics method out of the wide range of methods relies on a manual, heuristic trial-and-error process. We introduce a novel automated framework that optimizes radiomics workflow construction per application by standardizing the radiomics workflow in modular components, including a large collection of algorithms for each component, and formulating a combined algorithm selection and hyperparameter optimization problem. To solve it, we employ automated machine learning through two strategies (random search and Bayesian optimization) and three ensembling approaches. Results show that a medium-sized random search and straight-forward ensembling perform similar to more advanced methods while being more efficient. Validated across twelve clinical applications, our approach outperforms both a radiomics baseline and human experts. Concluding, our framework improves and streamlines radiomics research by fully automatically optimizing radiomics workflow construction. To facilitate reproducibility, we publicly release six datasets, software of the method, and code to reproduce this study.
Bayesian optimization (BO) is a flexible and powerful framework that is suitable for computationally expensive simulation-based applications and guarantees statistical convergence to the global optimum. While remaining as one of the most popular optimization methods, its capability is hindered by the size of data, the dimensionality of the considered problem, and the nature of sequential optimization. These scalability issues are intertwined with each other and must be tackled simultaneously. In this work, we propose the Scalable$^3$-BO framework, which employs sparse GP as the underlying surrogate model to scope with Big Data and is equipped with a random embedding to efficiently optimize high-dimensional problems with low effective dimensionality. The Scalable$^3$-BO framework is further leveraged with asynchronous parallelization feature, which fully exploits the computational resource on HPC within a computational budget. As a result, the proposed Scalable$^3$-BO framework is scalable in three independent perspectives: with respect to data size, dimensionality, and computational resource on HPC. The goal of this work is to push the frontiers of BO beyond its well-known scalability issues and minimize the wall-clock waiting time for optimizing high-dimensional computationally expensive applications. We demonstrate the capability of Scalable$^3$-BO with 1 million data points, 10,000-dimensional problems, with 20 concurrent workers in an HPC environment.
From ultra-sensitive detectors of fundamental forces to quantum networks and sensors, mechanical resonators are enabling next-generation technologies to operate in room temperature environments. Currently, silicon nitride nanoresonators stand as a leading microchip platform in these advances by allowing for mechanical resonators whose motion is remarkably isolated from ambient thermal noise. However, to date, human intuition has remained the driving force behind design processes. Here, inspired by nature and guided by machine learning, a spiderweb nanomechanical resonator is developed that exhibits vibration modes which are isolated from ambient thermal environments via a novel "torsional soft-clamping" mechanism discovered by the data-driven optimization algorithm. This bio-inspired resonator is then fabricated; experimentally confirming a new paradigm in mechanics with quality factors above 1 billion in room temperature environments. In contrast to other state-of-the-art resonators, this milestone is achieved with a compact design which does not require sub-micron lithographic features or complex phononic bandgaps, making it significantly easier and cheaper to manufacture at large scales. Here we demonstrate the ability of machine learning to work in tandem with human intuition to augment creative possibilities and uncover new strategies in computing and nanotechnology.