Acquisition Of Functions Research Articles

葵花种植机械电动化改造及控制系统设计

This research aimed to address the issue of physical damage caused by long-term work in high-saline fields by electrifying sunflower growing machinery and designing control systems. Firstly, the method of establishing the equation of motion is used to obtain the parameters to carry out motorization modification, and then the control system is built using STM32 as the main controller, which has the functions of control, data acquisition, and alarm. The feasibility and stability of the study are verified by using a model vehicle to conduct tests of straight-line driving and steering in a simulated high-saline field.

Multi-Objective BiLevel Optimization by Bayesian Optimization

In a multi-objective optimization problem, a decision maker has more than one objective to optimize. In a bilevel optimization problem, there are the following two decision-makers in a hierarchy: a leader who makes the first decision and a follower who reacts, each aiming to optimize their own objective. Many real-world decision-making processes have various objectives to optimize at the same time while considering how the decision-makers affect each other. When both features are combined, we have a multi-objective bilevel optimization problem, which arises in manufacturing, logistics, environmental economics, defence applications and many other areas. Many exact and approximation-based techniques have been proposed, but because of the intrinsic nonconvexity and conflicting multiple objectives, their computational cost is high. We propose a hybrid algorithm based on batch Bayesian optimization to approximate the upper-level Pareto-optimal solution set. We also extend our approach to handle uncertainty in the leader’s objectives via a hypervolume improvement-based acquisition function. Experiments show that our algorithm is more efficient than other current methods while successfully approximating Pareto-fronts.

Pareto Front-Diverse Batch Multi-Objective Bayesian Optimization

We consider the problem of multi-objective optimization (MOO) of expensive black-box functions with the goal of discovering high-quality and diverse Pareto fronts where we are allowed to evaluate a batch of inputs. This problem arises in many real-world applications including penicillin production where diversity of solutions is critical. We solve this problem in the framework of Bayesian optimization (BO) and propose a novel approach referred to as Pareto front-Diverse Batch Multi-Objective BO (PDBO). PDBO tackles two important challenges: 1) How to automatically select the best acquisition function in each BO iteration, and 2) How to select a diverse batch of inputs by considering multiple objectives. We propose principled solutions to address these two challenges. First, PDBO employs a multi-armed bandit approach to select one acquisition function from a given library. We solve a cheap MOO problem by assigning the selected acquisition function for each expensive objective function to obtain a candidate set of inputs for evaluation. Second, it utilizes Determinantal Point Processes (DPPs) to choose a Pareto-front-diverse batch of inputs for evaluation from the candidate set obtained from the first step. The key parameters for the methods behind these two steps are updated after each round of function evaluations. Experiments on multiple MOO benchmarks demonstrate that PDBO outperforms prior methods in terms of both the quality and diversity of Pareto solutions.

Constrained Bayesian Optimization under Partial Observations: Balanced Improvements and Provable Convergence

The partially observable constrained optimization problems (POCOPs) impede data-driven optimization techniques since an infeasible solution of POCOPs can provide little information about the objective as well as the constraints. We endeavor to design an efficient and provable method for expensive POCOPs under the framework of constrained Bayesian optimization. Our method consists of two key components. Firstly, we present an improved design of the acquisition functions that introduce balanced exploration during optimization. We rigorously study the convergence properties of this design to demonstrate its effectiveness. Secondly, we propose Gaussian processes embedding different likelihoods as the surrogate model for partially observable constraints. This model leads to a more accurate representation of the feasible regions compared to traditional classification-based models. Our proposed method is empirically studied on both synthetic and real-world problems. The results demonstrate the competitiveness of our method for solving POCOPs.

FlexiBO: A Decoupled Cost-Aware Multi-objective Optimization Approach for Deep Neural Networks (Abstract Reprint)

The design of machine learning systems often requires trading off different objectives, for example, prediction error and energy consumption for deep neural networks (DNNs). Typically, no single design performs well in all objectives; therefore, finding Pareto-optimal designs is of interest. The search for Pareto-optimal designs involves evaluating designs in an iterative process, and the measurements are used to evaluate an acquisition function that guides the search process. However, measuring different objectives incurs different costs. For example, the cost of measuring the prediction error of DNNs is orders of magnitude higher than that of measuring the energy consumption of a pre-trained DNN as it requires re-training the DNN. Current state-of-the-art methods do not consider this difference in objective evaluation cost, potentially incurring expensive evaluations of objective functions in the optimization process. In this paper, we develop a novel decoupled and cost-aware multi-objective optimization algorithm, which we call Flexible Multi-Objective Bayesian Optimization (FlexiBO) to address this issue. For evaluating each design, FlexiBO selects the objective with higher relative gain by weighting the improvement of the hypervolume of the Pareto region with the measurement cost of each objective. This strategy, therefore, balances the expense of collecting new information with the knowledge gained through objective evaluations, preventing FlexiBO from performing expensive measurements for little to no gain. We evaluate FlexiBO on seven state-of-the-art DNNs for image recognition, natural language processing (NLP), and speech-to-text translation. Our results indicate that, given the same total experimental budget, FlexiBO discovers designs with 4.8% to 12.4% lower hypervolume error than the best method in state-of-the-art multi-objective optimization.

Multi-Objective Bayesian Optimization with Active Preference Learning

There are a lot of real-world black-box optimization problems that need to optimize multiple criteria simultaneously. However, in a multi-objective optimization (MOO) problem, identifying the whole Pareto front requires the prohibitive search cost, while in many practical scenarios, the decision maker (DM) only needs a specific solution among the set of the Pareto optimal solutions. We propose a Bayesian optimization (BO) approach to identifying the most preferred solution in the MOO with expensive objective functions, in which a Bayesian preference model of the DM is adaptively estimated by an interactive manner based on the two types of supervisions called the pairwise preference and improvement request. To explore the most preferred solution, we define an acquisition function in which the uncertainty both in the objective function and the DM preference is incorporated. Further, to minimize the interaction cost with the DM, we also propose an active learning strategy for the preference estimation. We empirically demonstrate the effectiveness of our proposed method through the benchmark function optimization and the hyper-parameter optimization problems for machine learning models.

Abstract 4141: Quantitative measurement of 3D tumor spheroid invasion using live cell time-lapse imaging

Abstract 90% of the death of cancer individuals is related to cancer metastasis1. The invasion and migration of cancer cells into surrounding tissues and chambers is the first step of metastasis2. To have a better and more in-depth understanding of the pathophysiological activities involved in metastatic cancer, accurate and reliable methods for evaluating cell invasion are urgently needed. Compared to the 2D cell model, the 3D system mimics the natural physiological properties and conditions, such as structure, physiology, biological signals of living tissues, cell-matrix interactions4. In this study, we followed the commonly used protocol for 3D tumor invasion assay3 to generate spheroids in Ultra-Low-attachment (ULA) round bottom 96-well plate and subsequently embedded them in Matrigel® (5 mg/mL). After polymerization, the Matrigel that contains tumor spheroid was submerged in 100 μL cell culture media. The invasion of the 3D spheroid in the presence or absence of drug treatment was monitored and quantified using the brightfield and fluorescence live cell time-lapse image acquisition and image analysis functions of the Agilent xCELLigence RTCA eSight. The progress of spheroid invasion and the drug-inhibitory effects on tumor invasion were quantified by tracking and calculating the area of invasive protrusion from the spheroid. Our data show that 1) different types of cancer cells had different metastatic potentials demonstrated by the level of invadopodium extending into the Matrigel at desired time points; 2) the invasion of 3D HT1080 spheroid was inhibited by cytochalasin D (Cyto D), an inhibitor of actin polymerization, and GM6001, a potent broad-spectrum inhibitor of matrix metalloproteinases (MMP), in a dose-dependent manner; 3) Cyto D exhibited higher efficacy of inhibiting tumor invasion than GM6001. In summary, our approach by combining the 3D tumor spheroid invasion model with the RTCA eSight’s capability of live cell imaging and automatic quantification of the area of invasive protrusion can facilitate and enhance the development of new treatments at the preclinical stage in the future. Citation Format: Grace Yang, Tian Wang, Peifang Ye, Xiaoyu Zhang. Quantitative measurement of 3D tumor spheroid invasion using live cell time-lapse imaging [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4141.

Process quality control through Bayesian optimization with adaptive local convergence

How to take as few experiments as possible to achieve the desired quality target is a challenging and essential topic of process quality control, especially for complex and expensive manufacturing processes with high-dimensional quality outputs. This paper proposes a Bayesian optimization (BO)-based strategy utilizing quality information with adaptive local convergence. First, a series of pretreatment setting rules of the Gaussian process model is proposed to reduce the uncertainty of BO. Then, the quality target is adopted into the Gaussian process model, and the acquisition function is improved. Finally, to balance global exploration and local exploitation and reduce the experimental cost, an adaptive criterion is designed to switch global BO search to local search. The operating conditions can be efficiently updated until the quality target is reached. Two applications with high-dimensional quality outputs are presented to demonstrate the effectiveness of the proposed method.

Dynamic Exploration–Exploitation Pareto Approach for high-dimensional expensive black-box optimization

Surrogate Optimization (SO) plays a vital role in optimizing performance parameters for computationally expensive simulations. However, SO encounters significant challenges in high-dimensional spaces due to the curse of dimensionality, hampering effective point sampling around global optima. In this paper, we introduce “Dynamic Exploration–Exploitation Pareto Approach (DEEPA),” a novel SO method that combines Pareto sampling with a dynamic discretization schema to optimize high-dimensional black-box functions. Unlike traditional SO methods that heavily rely on specific surrogate models, Pareto sampling offers a more adaptable approach. Improvement-based acquisition functions, frequently employed in black-box optimization, are sensitive to model accuracy and tend to prioritize exploitation, potentially missing valuable regions of interest in complex landscapes. Furthermore, they can encounter challenges when dealing with high-dimensional problems due to the curse of dimensionality. DEEPA leverages dynamic coordinate importance to generate samples effectively, providing a solution for addressing high-dimensionality and complex functions. We employ feature selection strategies to assign importance probabilities to perturb each coordinate, demonstrating the impact of importance-based perturbation on convergence to a near-optimal region. We showcase DEEPA’s versatility in fixed-batch evaluation environments using complex global optimization test problems with various topological properties. We compare DEEPA’s performance with state-of-the-art black-box optimization algorithms, and our experimental results demonstrate DEEPA’s superior performance, particularly in complex problems with multiple local minima.

Bayesian Optimization via Exact Penalty

Constrained optimization problems pose challenges when the objective function and constraints are nonconvex and their evaluation requires expensive black-box simulations. Recently, hybrid optimization methods that integrate statistical surrogate modeling with numerical optimization algorithms have shown great promise, as they inherit the properties of global convergence from statistical surrogate modeling and fast local convergence from numerical optimization algorithms. However, the computational efficiency is not satisfied by practical needs under limited budgets and in the presence of equality constraints. In this article, we propose a novel hybrid optimization method, called exact penalty Bayesian optimization (EPBO), which employs Bayesian optimization within the exact penalty framework. We model the composite penalty function by a weighted sum of Gaussian processes, where the qualitative components of the constraint violations are smoothed by their predictive means. The proposed method features (i) closed-form acquisition functions, (ii) robustness to initial designs, (iii) the capability to start from infeasible points, and (iv) effective handling of equality constraints. We demonstrate the superiority of EPBO to state-of-the-art competitors using a suite of benchmark synthetic test problems and two real-world engineering design problems.

Bayesian optimization for geoacoustic inversion

Geoacoustic inversion of high-dimensional parameter spaces is a computationally intensive procedure, often necessitating thousands of forward model evaluations to accurately estimate the geoacoustic environment, such as Markov chain Monte Carlo sampling. This study introduces Bayesian optimization (BO), an efficient global optimization technique, to estimate geoacoustic parameters with significantly fewer evaluations, typically on the order of hundreds. BO involves an iterative search within the parameter space to locate the global optimum of an objective function; in this study, the Bartlett power is used. BO consists of fitting a Gaussian process surrogate model to existing evaluations of the objective function, followed by selecting a new data point for evaluation using a heuristic acquisition function. The effectiveness of BO is showcased through its application to both simulated and real-world data from a shallow-water environment for multidimensional parameter space encompassing source location, array tilt, and seabed properties.

Automated experimental design of safe rampdowns via probabilistic machine learning

Typically the rampdown phase of a shot consists of a decrease in current and injected power and optionally a change in shape, but there is considerable flexibility in the rate, sequencing, and duration of these changes. On the next generation of tokamaks it is essential that this is done safely as the device could be damaged by the stored thermal and electromagnetic energy present in the plasma. This works presents a procedure for automatically choosing experimental rampdown designs to rapidly converge to an effective rampdown trajectory. This procedure uses probabilistic machine learning methods paired with acquisition functions taken from Bayesian optimization. In a set of 2022 experiments at DIII-D, the rampdown designs produced by our method maintained plasma control down to substantially lower current and energy levels than are typically observed. The actions predicted by the model significantly improved as the model was able to explore over the course of the experimental campaign.

A Bayesian sampling optimisation strategy for finite element model updating

AbstractModel Updating (MU) aims to estimate the unknown properties of a physical system of interest from experimental observations. In Finite Element (FE) models, these unknowns are the elements’ parameters. Typically, besides model calibration purposes, MU and FEMU procedures are employed for the Non-Destructive Evaluation (NDE) and damage assessment of structures. In this framework, damage can be located and quantified by updating the parameters related to stiffness. However, these procedures require the minimisation of a cost function, defined according to the difference between the model and the experimental data. Sophisticated FE models can generate expensive and non-convex cost functions, which minimization is a non-trivial task. To deal with this challenging optimization problem, this work makes use of a Bayesian sampling optimisation technique. This approach consists of generating a statistical surrogate model of the underlying cost function (in this case, a Gaussian Process is used) and applying an acquisition function that drives the intelligent selection of the next sampling point, considering both exploitation and exploration needs. This results in a very efficient yet very powerful optimization technique, necessitating of minimal sampling volume. The performance of this proposed scheme is then compared to three well-established global optimisation algorithms. This investigation is performed on numerical and experimental case studies based on the famous Mirandola bell tower.

Model Change Active Learning in Graph-Based Semi-supervised Learning

Active learning in semi-supervised classification involves introducing additional labels for unlabelled data to improve the accuracy of the underlying classifier. A challenge is to identify which points to label to best improve performance while limiting the number of new labels. “Model Change” active learning quantifies the resulting change incurred in the classifier by introducing the additional label(s). We pair this idea with graph-based semi-supervised learning (SSL) methods, that use the spectrum of the graph Laplacian matrix, which can be truncated to avoid prohibitively large computational and storage costs. We consider a family of convex loss functions for which the acquisition function can be efficiently approximated using the Laplace approximation of the posterior distribution. We show a variety of multiclass examples that illustrate improved performance over prior state-of-art.

A stochastic data-driven Bayesian optimization approach for intensified ethanol–water separation systems

The optimization of advanced intensified distillation operation systems is crucial for achieving sustainability in the chemical industry. In this paper, we present a stochastic data-driven Bayesian optimization approach for the design and operation of ethanol distillation separation systems. The proposed optimal design methodology does not require a first-principles mathematical model to draw optimal operating conditions. The approach involves an initial sampling of input/output system measurements, followed by the construction of an approximate input/output Gaussian process model. The next best sampling point is then drawn using an acquisition function, and the algorithm converges when the optimization function value does not change or when the maximum number of iterations is reached. We address the optimal parametric design of advanced ethanol–water intensified distillation separation systems, assuming a mixture of the ethanol–water system is available, and almost anhydrous ethanol is required for mobility applications. We also consider the impact of noisy measurements on the optimality operating region. The proposed approach can be extended to the optimization of merged discrete and continuous systems. Bayesian optimization techniques represent black-box optimization strategies that can be employed for the continuous online optimization of intricate separation systems. These techniques aid in mitigating sustainability issues by reducing mass and energy requirements.

Sequential Bayesian-optimized graphene synthesis by direct solar-thermal chemical vapor deposition

This work reports the use of a high-flux solar simulator that mimics the solar spectrum and a cold-wall CVD reactor to demonstrate the feasibility of utilizing a renewable energy resource in synthesizing graphene under various conditions. A parametric study of process parameters was conducted using a probabilistic approach. Gaussian process regression serves as a surrogate to establish a prior for Bayesian optimization, and an information acquisition function is employed to identify conditions that yield high-quality products. Backscattered electron images and Raman mapping were used to assess the effects of growth conditions on graphene characteristic sizes, film quality, and uniformity. We report the synthesis of high-quality single-layer graphene (SLG) and AB-stacked bilayer graphene films in a one-step, short-time process with ID/IG\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$I_{D}/I_{G}$$\\end{document} ratios of 0.21 and 0.14, respectively. Electron diffraction analysis shows peak intensities that resemble SLG and AB-bilayer graphene with up to 5 and 20 μ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${\\upmu }$$\\end{document}m grain sizes, respectively. The optical transmissivities of SLG and AB-bilayer graphene fall between 0.959–0.977 and 0.929–0.953, whereas the sheet resistances measured by a 4-point probe with 1 mm spacing are 15.5 ± 4.6 and 3.4 ± 1.5 kΩ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\Omega$$\\end{document}/sq, respectively. Further scale-up of the optimized graphene growth area was achieved by flattening the insolation profile, leading to spatial uniformity up to 13 mm in radius. Direct solar capture for CVD synthesis enable a practical and sustainable option for synthesizing graphene films applicable for photonic and electronic applications.

Human-in-the-Loop: The Future of Machine Learning in Automated Electron Microscopy

Abstract Machine learning (ML) methods are progressively gaining acceptance in the electron microscopy community for de-noising, semantic segmentation, and dimensionality reduction of data post-acquisition. The introduction of the application programming interfaces (APIs) by major instrument manufacturers now allows the deployment of ML workflows in microscopes, not only for data analytics but also for real-time decision-making and feedback for microscope operation. However, the number of use cases for real-time ML remains remarkably small. Here, we discuss some considerations in designing ML-based active experiments and pose that the likely strategy for the next several years will be human-in-the-loop automated experiments (hAE). In this paradigm, the ML learning agent directly controls beam position and image and spectroscopy acquisition functions, and a human operator monitors experiment progression in real and feature space of the system and tunes the policies of the ML agent to steer the experiment toward specific objectives.

Eight-channel signal synchronous acquisition system based on PCI and LabVIEW

In response to the multi-channel acquisition of the power grid simulator data collection system, this article builds a 8-channel signal synchronization acquisition system based on PCI-1706U and LabView, which introduces the characteristics of PCI-1706U hardware and LabView software DAQ system. A LabView DAQ program that uses the PCI-1706U to collect 8-way signal from the signal generator, complete the experiments of multi-channel synchronous acquisition and delayed start to collection, verify the feasibility and stability of the system, and analyzed the collected signals. Finally, the main work and innovation points of the thesis are summarized, and suggestions and outlook for future improvement are given. This study provides an effective implementation method based on PCI-1706U and LabView multi-channel signal synchronous acquisition system and delay acquisition function, with certain application value and promotion space.

Bayesian Optimisation for Constrained Problems

Many real-world optimisation problems such as hyperparameter tuning in machine learning or simulation-based optimisation can be formulated as expensive-to-evaluate black-box functions. A popular approach to tackle such problems is Bayesian optimisation, which builds a response surface model based on the data collected so far, and uses the mean and uncertainty predicted by the model to decide what information to collect next. In this paper, we propose a generalisation of the well-known Knowledge Gradient acquisition function that allows it to handle constraints. We empirically compare the new algorithm with four other state-of-the-art constrained Bayesian optimisation algorithms and demonstrate its superior performance. We also prove theoretical convergence in the infinite budget limit.

Deep learning method for predicting electromagnetic emission spectrum of aerospace equipment

AbstractThis paper proposes a deep learning method to predict the electromagnetic emission spectrum in the electromagnetic compatibility (EMC) test of aerospace products. A threshold‐based data decomposition method is used to propose the spike signal, reconstruct the original test data, and solve the contradiction between the overfitting and prediction accuracy of the deep learning method to deal with the EMC test spectrum. Using a long short‐term memory neural network architecture for predicting electromagnetic emission spectrum, the Bayesian optimization method is used to optimize the network hyperparameter, and the acquisition function is introduced into the loss function to improve the comprehensive training optimization of deep learning network. Apply the method to three numerical examples: signal line current conduction emission, power line voltage conduction emission, and electric field radiation emission. The analysis results indicate that at a 95% confidence level, the predicted electromagnetic emission spectrum is basically consistent with the test results. The prediction results can be used as the basis for EMC evaluation of aerospace electronic equipment.