Multifidelity and Multiscale Bayesian Framework for High-Dimensional Engineering Design and Calibration

2019 ◽  
Vol 141 (12) ◽  
Author(s):  
Soumalya Sarkar ◽  
Sudeepta Mondal ◽  
Michael Joly ◽  
Matthew E. Lynch ◽  
Shaunak D. Bopardikar ◽  
...  
Keyword(s):  
Machine Learning ◽  
Learning Strategies ◽  
Optimization Problems ◽  
Bayesian Optimization ◽  
High Dimensional ◽  
Optimization Approach ◽  
Information Theoretic ◽  
Compressor Rotor ◽  
Practical Engineering ◽  
Microstructure Prediction

AbstractThis paper proposes a machine learning–based multifidelity modeling (MFM) and information-theoretic Bayesian optimization approach where the associated models can have complex discrepancies among each other. Advantages of MFM-based optimization over a single-fidelity surrogate, specifically under complex constraints, are discussed with benchmark optimization problems involving noisy data. The MFM framework, based on modeling of the varied fidelity information sources via Gaussian processes, is augmented with information-theoretic active learning strategies that involve sequential selection of optimal points in a multiscale architecture. This framework is demonstrated to exhibit improved efficiency on practical engineering problems like high-dimensional design optimization of compressor rotor via implementing its multiscale architecture and calibration of expensive microstructure prediction model. From the perspective of the machine learning–assisted design of multiphysics systems, advantages of the proposed framework have been presented with respect to accelerating the search of optimal design conditions under budget constraints.

Machine Learning to Aid Tuning of Numerical Parameters in Topology Optimization

2019 ◽  
Vol 141 (11) ◽  
Author(s):  
Matthew E. Lynch ◽  
Soumalya Sarkar ◽  
Kurt Maute
Keyword(s):  
Machine Learning ◽  
Topology Optimization ◽  
Optimization Problems ◽  
Bayesian Optimization ◽  
Computational Time ◽  
Optimization Approach ◽  
Learning Approaches ◽  
Solution Quality ◽  
Topology Optimization Problem ◽  
Tuning Parameters

Abstract Recent advances in design optimization have significant potential to improve the function of mechanical components and systems. Coupled with additive manufacturing, topology optimization is one category of numerical methods used to produce algorithmically generated optimized designs making a difference in the mechanical design of hardware currently being introduced to the market. Unfortunately, many of these algorithms require extensive manual setup and control, particularly of tuning parameters that control algorithmic function and convergence. This paper introduces a framework based on machine learning approaches to recommend tuning parameters to a user in order to avoid costly trial and error involved in manual tuning. The algorithm reads tuning parameters from a repository of prior, similar problems adjudged using a dissimilarity metric based on problem metadata and refines them for the current problem using a Bayesian optimization approach. The approach is demonstrated for a simple topology optimization problem with the objective of achieving good topology optimization solution quality and then with the additional objective of finding an optimal “trade” between solution quality and required computational time. The goal is to reduce the total number of “wasted” tuning runs that would be required for purely manual tuning. With more development, the framework may ultimately be useful on an enterprise level for analysis and optimization problems—topology optimization is one example but the framework is also applicable to other optimization problems such as shape and sizing and in high-fidelity physics-based analysis models—and enable these types of advanced approaches to be used more efficiently.

scholarly journals Lost in Optimization of Water Distribution Systems: Better Call Bayes

2022 ◽  
Author(s):  
Antonio Candelieri ◽  
Andrea Ponti ◽  
Ilaria Giordani ◽  
Francesco Archetti
Keyword(s):  
Distribution Systems ◽  
Water Distribution ◽  
Probabilistic Models ◽  
Optimization Problems ◽  
Simulation Optimization ◽  
Water Distribution Systems ◽  
Bayesian Optimization ◽  
Optimization Approach ◽  
Hydraulic Simulation ◽  
Multi Objective

The main goal of this paper is to show that Bayesian optimization could be regarded as a general framework for the data driven modelling and solution of problems arising in water distribution systems. Hydraulic simulation, both scenario based, and Monte Carlo is a key tool in modelling in water distribution systems. The related optimization problems fall in a simulation/optimization framework in which objectives and constraints are often black-box. Bayesian Optimization (BO) is characterized by a surrogate model, usually a Gaussian process, but also a random forest and increasingly neural networks and an acquisition function which drives the search for new evaluation points. These modelling options make BO nonparametric, robust, flexible and sample efficient particularly suitable for simulation/optimization problems. A defining characteristic of BO is its versatility and flexibility, given for instance by different probabilistic models, in particular different kernels, different acquisition functions. These characteristics of the Bayesian optimization approach are exemplified by the two problems: cost/energy optimization in pump scheduling and optimal sensor placement for early detection on contaminant intrusion. Different surrogate models have been used both in explicit and implicit control schemes. Showing that BO can drive the process of learning control rules directly from operational data. BO can also be extended to multi-objective optimization. Two algorithms have been proposed for multi-objective detection problem using two different acquisition functions.

Robust Optimization With NURBS HyPerModels

2008 ◽  
Author(s):  
Abiola M. Ajetunmobi ◽  
Cameron J. Turner ◽  
Richard H. Crawford
Keyword(s):  
Robust Optimization ◽  
Optimization Problems ◽  
Control Point ◽  
Optimization Methods ◽  
Design System ◽  
Optimization Approach ◽  
Engineering Systems ◽  
Parameter Uncertainties ◽  
Real Time System ◽  
Practical Engineering

Engineering systems are generally susceptible to parameter uncertainties that influence real-time system performance and long-term system reliability. However, designers and engineers must design system solutions that are both optimal and dependable. Robust design techniques and robust optimization methods in particular, have emerged as promising methodologies to address the problem of dealing with parameter uncertainties. This research advances a robust optimization approach that exploits gradient information embedded in proximate NURBs control point clusters that are inherent in NURBs metamodel design space representations. The proximate control point clusters embody the target sensitivity profile and therefore include robust optimal solutions, thus enabling selective optimization within regions associated with the clusters. This robust optimization framework has been implemented and is demonstrated on unconstrained robust optimization problems from two test functions and a constrained robust optimization problem from a practical engineering design problem.

Machine learning-aided cost prediction and optimization in construction operations

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Virok Sharma ◽  
Mohd Zaki ◽  
Kumar Neeraj Jha ◽  
N. M. Anoop Krishnan
Keyword(s):  
Machine Learning ◽  
Environmental Impact ◽  
Gaussian Process Regression ◽  
Construction Cost ◽  
Bayesian Optimization ◽  
Optimization Approach ◽  
Cost Prediction ◽  
Construction Operations ◽  
Content Type ◽  
Multi Objective

PurposeThis paper aims to use a data-driven approach towards optimizing construction operations. To this extent, it presents a machine learning (ML)-aided optimization approach, wherein the construction cost is predicted as a function of time, resources and environmental impact, which is further used as a surrogate model for cost optimization.Design/methodology/approachTaking a dataset from literature, the paper has applied various ML algorithms, namely, simple and regularized linear regression, random forest, gradient boosted trees, neural network and Gaussian process regression (GPR) to predict the construction cost as a function of time, resources and environmental impact. Further, the trained models were used to optimize the construction cost applying single-objective (with and without constraints) and multi-objective optimizations, employing Bayesian optimization, particle swarm optimization (PSO) and non-dominated sorted genetic algorithm.FindingsThe results presented in the paper demonstrate that the ensemble methods, such as gradient boosted trees, exhibit the best performance for construction cost prediction. Further, it shows that multi-objective optimization can be used to develop a Pareto front for two competing variables, such as cost and environmental impact, which directly allows a practitioner to make a rational decision.Research limitations/implicationsNote that the sequential nature of events which dictates the scheduling is not considered in the present work. This aspect could be incorporated in the future to develop a robust scheme that can optimize the scheduling dynamically.Originality/valueThe paper demonstrates that a ML approach coupled with optimization could enable the development of an efficient and economic strategy to plan the construction operations.

scholarly journals A New Multi-Objective Bayesian Optimization Formulation With the Acquisition Function for Convergence and Diversity

2020 ◽  
Vol 142 (9) ◽  
Author(s):  
Leshi Shu ◽  
Ping Jiang ◽  
Xinyu Shao ◽  
Yan Wang
Keyword(s):  
Optimization Problems ◽  
Computational Cost ◽  
Bayesian Optimization ◽  
Optimization Approach ◽  
Exploration And Exploitation ◽  
Pareto Solutions ◽  
Multi Objective ◽  
Trade Offs ◽  
Conflicting Objectives ◽  
Optimization Formulation

Abstract Bayesian optimization is a metamodel-based global optimization approach that can balance between exploration and exploitation. It has been widely used to solve single-objective optimization problems. In engineering design, making trade-offs between multiple conflicting objectives is common. In this work, a multi-objective Bayesian optimization approach is proposed to obtain the Pareto solutions. A novel acquisition function is proposed to determine the next sample point, which helps improve the diversity and convergence of the Pareto solutions. The proposed approach is compared with some state-of-the-art metamodel-based multi-objective optimization approaches with four numerical examples and one engineering case. The results show that the proposed approach can obtain satisfactory Pareto solutions with significantly reduced computational cost.

scholarly journals Multilabelled Optimal Feature Classification Procedure for High Dimensional Bio Medical Data

2019 ◽  
Vol 8 (12) ◽  
pp. 262-266
Keyword(s):  
Machine Learning ◽  
Medical Knowledge ◽  
Medical Data ◽  
Medical Decision ◽  
High Dimensional ◽  
Optimization Approach ◽  
Protein Protein Interaction ◽  
Medical Decision Support ◽  
Interaction Extraction ◽  
Optimal Feature

The Machine Learning field has extended its thrust virtually in any domain of analysis and within the near past has become a trusted tool in the medical domain. The experiential domain of automatic learning is employed in tasks like medical decision support, medical imaging, protein-protein interaction, extraction of medical data, and for overall patient management care. ML is pictured as a tool by that computer-based systems are often integrated within the health care field so as to induce a far better, well-organized treatment. To extract optimal feature selection with high dimensional bio-medical knowledge, during this paper propose a Advance Machine Learning Approach with optimization approach i.e. Ant Colony Optimization (ACO). It extracts sentences from revealed medical papers that mention diseases and coverings, and identifies semantic relations that exit between diseases and coverings. Our analysis results for these tasks show that the projected methodology obtains reliable outcomes that might be integrated in associative application to be employed in the treatment domain.

Supervised Classifier Approach for Intrusion Detection on KDD with Optimal MapReduce Framework Model in Cloud Computing

2019 ◽  
Vol 12 ◽  
Author(s):  
M. Ilayaraja ◽  
S. Hemalatha ◽  
P. Manickam ◽  
K. Sathesh Kumar ◽  
K. Shankar
Keyword(s):  
Machine Learning ◽  
Cloud Computing ◽  
Intrusion Detection ◽  
Decision Tree ◽  
Learning Strategies ◽  
Nearest Neighbor ◽  
Detection System ◽  
K Nearest Neighbor ◽  
Mapreduce Model ◽  
The Web

Cloud computing is characterized as the arrangement of assets or administrations accessible through the web to the clients on their request by cloud providers. It communicates everything as administrations over the web in view of the client request, for example operating system, organize equipment, storage, assets, and software. Nowadays, Intrusion Detection System (IDS) plays a powerful system, which deals with the influence of experts to get actions when the system is hacked under some intrusions. Most intrusion detection frameworks are created in light of machine learning strategies. Since the datasets, this utilized as a part of intrusion detection is Knowledge Discovery in Database (KDD). In this paper detect or classify the intruded data utilizing Machine Learning (ML) with the MapReduce model. The primary face considers Hadoop MapReduce model to reduce the extent of database ideal weight decided for reducer model and second stage utilizing Decision Tree (DT) classifier to detect the data. This DT classifier comprises utilizing an appropriate classifier to decide the class labels for the non-homogeneous leaf nodes. The decision tree fragment gives a coarse section profile while the leaf level classifier can give data about the qualities that influence the label inside a portion. From the proposed result accuracy for detection is 96.21% contrasted with existing classifiers, for example, Neural Network (NN), Naive Bayes (NB) and K Nearest Neighbor (KNN).

scholarly journals Classification of Brainwaves for Sleep Stages by High-Dimensional FFT Features from EEG Signals

2020 ◽  
Vol 10 (5) ◽  
pp. 1797 ◽  
Author(s):  
Mera Kartika Delimayanti ◽  
Bedy Purnama ◽  
Ngoc Giang Nguyen ◽  
Mohammad Reza Faisal ◽  
Kunti Robiatul Mahmudah ◽  
...  
Keyword(s):  
Machine Learning ◽  
Sleep Stage ◽  
Machine Learning Algorithms ◽  
High Dimensional ◽  
Sleep Stages ◽  
Eeg Signals ◽  
Stage Classification ◽  
Sleep Stage Classification ◽  
Low Dimensional

Manual classification of sleep stage is a time-consuming but necessary step in the diagnosis and treatment of sleep disorders, and its automation has been an area of active study. The previous works have shown that low dimensional fast Fourier transform (FFT) features and many machine learning algorithms have been applied. In this paper, we demonstrate utilization of features extracted from EEG signals via FFT to improve the performance of automated sleep stage classification through machine learning methods. Unlike previous works using FFT, we incorporated thousands of FFT features in order to classify the sleep stages into 2–6 classes. Using the expanded version of Sleep-EDF dataset with 61 recordings, our method outperformed other state-of-the art methods. This result indicates that high dimensional FFT features in combination with a simple feature selection is effective for the improvement of automated sleep stage classification.

A microbial treatment population dynamics optimization approach

2021 ◽  
pp. 1-15
Author(s):  
Jinding Gao
Keyword(s):  
Population Dynamics ◽  
Information Exchange ◽  
Information Diffusion ◽  
Optimization Problems ◽  
Microbial Control ◽  
Function Optimization ◽  
Optimization Approach ◽  
Dynamics Model ◽  
Population Dynamics Model ◽  
Number Of Individuals

In order to solve some function optimization problems, Population Dynamics Optimization Algorithm under Microbial Control in Contaminated Environment (PDO-MCCE) is proposed by adopting a population dynamics model with microbial treatment in a polluted environment. In this algorithm, individuals are automatically divided into normal populations and mutant populations. The number of individuals in each category is automatically calculated and adjusted according to the population dynamics model, it solves the problem of artificially determining the number of individuals. There are 7 operators in the algorithm, they realize the information exchange between individuals the information exchange within and between populations, the information diffusion of strong individuals and the transmission of environmental information are realized to individuals, the number of individuals are increased or decreased to ensure that the algorithm has global convergence. The periodic increase of the number of individuals in the mutant population can greatly increase the probability of the search jumping out of the local optimal solution trap. In the iterative calculation, the algorithm only deals with 3/500∼1/10 of the number of individual features at a time, the time complexity is reduced greatly. In order to assess the scalability, efficiency and robustness of the proposed algorithm, the experiments have been carried out on realistic, synthetic and random benchmarks with different dimensions. The test case shows that the PDO-MCCE algorithm has better performance and is suitable for solving some optimization problems with higher dimensions.

Sign in / Sign up

Export Citation Format

Share Document