Nonpenalty Machine Learning Constraint Handling Using PSO-SVM for Structural Optimization

Firstly formulated to solve unconstrained optimization problems, the common way to solve constrained ones with the metaheuristic particle swarm optimization algorithm (PSO) is represented by adopting some penalty functions. In this paper, a new nonpenalty-based constraint handling approach for PSO is implemented, adopting a supervised classification machine learning method, the support vector machine (SVM). Because of its generality, constraint handling with SVM appears more adaptive both to nonlinear and discontinuous boundary. To preserve the feasibility of the population, a simple bisection algorithm is also implemented. To improve the search performances of the algorithm, a relaxation function of the constraints is also adopted. In the end part of this paper, two numerical literature benchmark examples and two structural examples are discussed. The first structural example refers to a homogeneous constant cross-section simply supported beam. The second one refers to the optimization of a plane simply supported Warren truss beam. The obtained results in terms of objective function demonstrate that this new approach represents a valid alternative to solve constrained optimization problems even in structural optimization field. Furthermore, as demonstrated by the Warren truss beam problem, this new algorithm provides an optimal structural design which represents also a good solution from the technical point of view with a trivial rounding-off that does not jeopardize significantly the optimization design process.

Download Full-text

Practical CO2—WAG Field Operational Designs Using Hybrid Numerical-Machine-Learning Approaches

Energies ◽

10.3390/en14041055 ◽

2021 ◽

Vol 14 (4) ◽

pp. 1055

Author(s):

Qian Sun ◽

William Ampomah ◽

Junyu You ◽

Martha Cather ◽

Robert Balch

Keyword(s):

Machine Learning ◽

Oil Recovery ◽

History Matching ◽

Optimization Problems ◽

Learning Technologies ◽

Petroleum Engineering ◽

Support Vector ◽

Learning Approaches ◽

Field Development ◽

Proxy Models

Machine-learning technologies have exhibited robust competences in solving many petroleum engineering problems. The accurate predictivity and fast computational speed enable a large volume of time-consuming engineering processes such as history-matching and field development optimization. The Southwest Regional Partnership on Carbon Sequestration (SWP) project desires rigorous history-matching and multi-objective optimization processes, which fits the superiorities of the machine-learning approaches. Although the machine-learning proxy models are trained and validated before imposing to solve practical problems, the error margin would essentially introduce uncertainties to the results. In this paper, a hybrid numerical machine-learning workflow solving various optimization problems is presented. By coupling the expert machine-learning proxies with a global optimizer, the workflow successfully solves the history-matching and CO2 water alternative gas (WAG) design problem with low computational overheads. The history-matching work considers the heterogeneities of multiphase relative characteristics, and the CO2-WAG injection design takes multiple techno-economic objective functions into accounts. This work trained an expert response surface, a support vector machine, and a multi-layer neural network as proxy models to effectively learn the high-dimensional nonlinear data structure. The proposed workflow suggests revisiting the high-fidelity numerical simulator for validation purposes. The experience gained from this work would provide valuable guiding insights to similar CO2 enhanced oil recovery (EOR) projects.

Download Full-text

Application of Particle Swarm Optimization Based on Support Vector Machine in Multi-Objective Structure Optimization

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.201-202.283 ◽

2012 ◽

Vol 201-202 ◽

pp. 283-286

Author(s):

Chen Yang Chang ◽

Jing Mei Zhai ◽

Qin Xiang Xia ◽

Bin Cai

Keyword(s):

Support Vector Machine ◽

Particle Swarm Optimization ◽

Optimization Design ◽

Optimization Problems ◽

Particle Swarm ◽

Structure Optimization ◽

Design Parameters ◽

Support Vector ◽

Swarm Optimization ◽

Multi Objective

Aiming at addressing optimization problems of complex mathematical model with large amount of calculation, a method based on support vector machine and particle swarm optimization for structure optimization design was proposed. Support Vector Machine (SVM) is a powerful computational tool for problems with nonlinearity and could establish approximate structures model. Grey relational analysis was utilized to calculate the coefficient between target parameters in order to change the multi-objective optimization problem into a single objective one. The reconstructed models were solved by Particle Swam Optimization (PSO) algorithm. A slip cover at medical treatment was adopted as an example to illustrate this methodology. Appropriate design parameters were selected through the orthogonal experiment combined with ANSYS. The results show this methodology is accurate and feasible, which provides an effective strategy to solve complex optimization problems.

Download Full-text

Classifying bridges for the risk of fire hazard via competitive machine learning

Advances in Bridge Engineering ◽

10.1186/s43251-020-00027-2 ◽

2021 ◽

Vol 2 (1) ◽

Author(s):

V. K. Kodur ◽

M. Z. Naser

Keyword(s):

Machine Learning ◽

Structural Engineering ◽

Fire Hazard ◽

Additive Model ◽

Experimental Tests ◽

Point Of View ◽

Concrete Bridges ◽

Hazard Evaluation ◽

Support Vector ◽

Analytical Approaches

AbstractThis study presents a machine learning (ML) approach to identify vulnerability of bridges to fire hazard. For developing this ML approach, data on a series of bridge fires was first collected and then analyzed through three algorithms; Random forest (RF), Support vector machine (SVM) and Generalize additive model (GAM), competing to yield the highest accuracy. As part of this analysis, 80 steel bridges and 38 concrete bridges were assessed. The outcome of this analysis shows that the ML based proposed approach can be effectively applied to arrive at the risk based classification of bridges from a fire hazard point of view. In addition, the developed ML algorithms are also capable of identifying the most critical features that govern bridges vulnerability to fire hazard. In parallel, this study showcases the potential of integrating ML into structural engineering applications as a supporting tool for analysis (i.e. in lieu of experimental tests, advanced simulations, and analytical approaches). This work emphasizes the need to compile data on bridge fires from around the world into a centralized and open source database to accelerate the integration of ML in to fire hazard evaluation.

Download Full-text

Constraint Handling in Bayesian Optimization -- A Comparative Study of Support Vector Machine, Augmented Lagrangian and Expected Feasible Improvement

10.1115/gt2021-58562 ◽

2021 ◽

Author(s):

Yuan Jin ◽

Zheyi Yang ◽

Shiran Dai ◽

Yann Lebret ◽

Olivier Jung

Keyword(s):

Support Vector Machine ◽

Comparative Study ◽

Constrained Optimization ◽

Augmented Lagrangian ◽

Optimization Problems ◽

Bayesian Optimization ◽

Constraint Handling ◽

Sequential Optimization ◽

Support Vector ◽

Complex Constraints

Abstract Many engineering problems involve complex constraints which can be computationally costly. To reduce the overall numerical cost, such constrained optimization problems are solved via surrogate models constructed on a Design of Experiment (DoE). Meanwhile, complex constraints may lead to infeasible initial DoE, which can be problematic for subsequent sequential optimization. In this study, we address constrained optimization problem in a Bayesian optimization framework. A comparative study is conducted to evaluate the performance of three approaches namely Expected Feasible Improvement (EFI) and slack Augmented Lagrangian method (AL) and Expected Improvement with Probabilistic Support Vector Machine in constraint handling with feasible or infeasible initial DoE. AL is capable to start sequential optimization with infeasible initial DoE, while EFI requires extra a priori enrichment to find at least one feasible sample. Empirical experiments are performed on both analytical functions and a low pressure turbine disc design problem. Through these benchmark problems, EFI and AL are shown to have overall similar performance in problems with inequality constraints. However, the performance of EIPSVM is affected strongly by the corresponding hyperparameter values. In addition, we show evidences that with an appropriate handling of infeasible initial DoE, EFI does not necessarily underperform compared with AL solving optimization problems with mixed inequality and equality constraints.

Download Full-text

The Performance of Penalty Methods on Tree-Seed Algorithm for Numerical Constrained Optimization Problems

The International Arab Journal of Information Technology ◽

10.34028/iajit/17/5/13 ◽

2020 ◽

Vol 17 (5) ◽

pp. 799-807 ◽

Cited By ~ 1

Author(s):

Ahmet Cinar ◽

Mustafa Kiran

Keyword(s):

Constrained Optimization ◽

Optimization Problems ◽

Constraint Handling ◽

Constrained Optimization Problems ◽

Swarm Optimization ◽

Bee Colony ◽

Unconstrained Optimization Problems ◽

Handling Methods ◽

State Of Art ◽

Tree Seed

The constraints are the most important part of many optimization problems. The metaheuristic algorithms are designed for solving continuous unconstrained optimization problems initially. The constraint handling methods are integrated into these algorithms for solving constrained optimization problems. Penalty approaches are not only the simplest way but also as effective as other constraint handling techniques. In literature, there are many penalty approaches and these are grouped as static, dynamic and adaptive. In this study, we collect them and discuss the key benefits and drawbacks of these techniques. Tree-Seed Algorithm (TSA) is a recently developed metaheuristic algorithm, and in this study, nine different penalty approaches are integrated with the TSA. The performance of these approaches is analyzed on well-known thirteen constrained benchmark functions. The obtained results are compared with state-of-art algorithms like Differential Evolution (DE), Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC), and Genetic Algorithm (GA). The experimental results and comparisons show that TSA outperformed all of them on these benchmark functions

Download Full-text

RECENT ADVANCES ON SUPPORT VECTOR MACHINES RESEARCH

Technological and Economic Development of Economy ◽

10.3846/20294913.2012.661205 ◽

2012 ◽

Vol 18 (1) ◽

pp. 5-33 ◽

Cited By ~ 86

Author(s):

Yingjie Tian ◽

Yong Shi ◽

Xiaohui Liu

Keyword(s):

Machine Learning ◽

Support Vector Machines ◽

Optimization Problems ◽

Optimization Methods ◽

Bankruptcy Prediction ◽

Optimization Models ◽

Support Vector ◽

Problem Solution ◽

Vector Machines ◽

Credit Risk Analysis

Support vector machines (SVMs), with their roots in Statistical Learning Theory (SLT) and optimization methods, have become powerful tools for problem solution in machine learning. SVMs reduce most machine learning problems to optimization problems and optimization lies at the heart of SVMs. Lots of SVM algorithms involve solving not only convex problems, such as linear programming, quadratic programming, second order cone programming, semi-definite programming, but also non-convex and more general optimization problems, such as integer programming, semi-infinite programming, bi-level programming and so on. The purpose of this paper is to understand SVM from the optimization point of view, review several representative optimization models in SVMs, their applications in economics, in order to promote the research interests in both optimization-based SVMs theory and economics applications. This paper starts with summarizing and explaining the nature of SVMs. It then proceeds to discuss optimization models for SVM following three major themes. First, least squares SVM, twin SVM, AUC Maximizing SVM, and fuzzy SVM are discussed for standard problems. Second, support vector ordinal machine, semisupervised SVM, Universum SVM, robust SVM, knowledge based SVM and multi-instance SVM are then presented for nonstandard problems. Third, we explore other important issues such as lp-norm SVM for feature selection, LOOSVM based on minimizing LOO error bound, probabilistic outputs for SVM, and rule extraction from SVM. At last, several applications of SVMs to financial forecasting, bankruptcy prediction, credit risk analysis are introduced.

Download Full-text

Improved Differential Evolution Algorithm for Structural Optimization Design of Flumes with Hybrid Discrete Variables

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.383-390.672 ◽

2011 ◽

Vol 383-390 ◽

pp. 672-677

Author(s):

Juan Zhou ◽

Duo Xin Zhang ◽

Xian Liang Liu

Keyword(s):

Structural Optimization ◽

Differential Evolution ◽

Optimization Design ◽

Optimization Problems ◽

Logistic Map ◽

Differential Evolution Algorithm ◽

Local Optimum ◽

Original Design ◽

Design Scheme ◽

De Algorithm

The traditional method applying to solve continuous variable optimization problems is not suit for flume structural optimization design with hybrid discrete variable. According to the mathematical model of structural optimum design of the prestressed U-shell flumes, differential evolution (DE) algorithm was introduced to flume structural optimization design. In order to improve the population’s diversity and the ability of escaping from the local optimum, a self-adapting crossover probability factor was presented. Furthermore, a chaotic sequence based on logistic map was employed to self-adaptively adjust mutation factor based on linear crossover, which can improve the convergence of DE algorithm. Dynamic penalty function, to transform the constrained problem to unconstrained one, was employed. The result shows that, compared with the original design scheme, the optimization design scheme can greatly reduce the amount of prestressed reinforcement. The construction cost of both the flume and the whole project can be reduced accordingly.

Download Full-text

Machine Learning Approach to Develop a Novel Multi-Objective Optimization Method for Pavement Material Proportion

Applied Sciences ◽

10.3390/app11020835 ◽

2021 ◽

Vol 11 (2) ◽

pp. 835 ◽

Cited By ~ 1

Author(s):

Chunyu Liang ◽

Xin Xu ◽

Heping Chen ◽

Wensheng Wang ◽

Kunkun Zheng ◽

...

Keyword(s):

Machine Learning ◽

Particle Swarm Optimization ◽

Optimization Design ◽

Optimization Problems ◽

Asphalt Pavement ◽

Asphalt Mixture ◽

Particle Swarm ◽

Computational Time ◽

Swarm Optimization ◽

Multi Objective

Asphalt mixture proportion design is one of the most important steps in asphalt pavement design and application. This study proposes a novel multi-objective particle swarm optimization (MOPSO) algorithm employing the Gaussian process regression (GPR)-based machine learning (ML) method for multi-variable, multi-level optimization problems with multiple constraints. First, the GPR-based ML method is proposed to model the objective and constraint functions without the explicit relationships between variables and objectives. In the optimization step, the metaheuristic algorithm based on adaptive weight multi-objective particle swarm optimization (AWMOPSO) is used to achieve the global optimal solution, which is very efficient for the objectives and constraints without mathematical relationships. The results showed that the optimal GPR model could describe the relationship between variables and objectives well in terms of root-mean-square error (RMSE) and R2. After the optimization by the proposed GPR-AWMOPSO algorithm, the comprehensive pavement performances were enhanced in terms of the permanent deformation resistance at high temperature, crack resistance at low temperature as well as moisture stability. Therefore, the proposed GPR-AWMOPSO algorithm is the best option and efficient for maximizing the performances of composite modified asphalt mixture. The GPR-AWMOPSO algorithm has advantages of less computational time and fewer samples, higher accuracy, etc. over traditional laboratory-based experimental methods, which can serve as guidance for the proportion optimization design of asphalt pavement.

Download Full-text

Machine Learning for Mass Spectrometry Data Analysis in Proteomics

Current Proteomics ◽

10.2174/1570164617999201023145304 ◽

2020 ◽

Vol 17 ◽

Author(s):

Juntao Li ◽

Kanglei Zhou ◽

Bingyu Mu

Keyword(s):

Machine Learning ◽

Mass Spectrometry ◽

Support Vector Machine ◽

Data Analysis ◽

Rapid Development ◽

Point Of View ◽

Mass Spectrometry Data ◽

Support Vector ◽

Learning Methods ◽

Machine Learning Methods

: With the rapid development of high-throughput techniques, mass spectrometry has been widely used for largescale protein analysis. To search for the existing proteins, discover biomarkers, and diagnose and prognose diseases, machine learning methods are applied in mass spectrometry data analysis. This paper reviews the applications of five kinds of machine learning methods to mass spectrometry data analysis from an algorithmic point of view, including support vector machine, decision tree, random forest, naive Bayesian classifier and deep learning.

Download Full-text

APPLYING ECONOMIC MEASURES TO LAPSE RISK MANAGEMENT WITH MACHINE LEARNING APPROACHES

Astin Bulletin ◽

10.1017/asb.2021.10 ◽

2021 ◽

pp. 1-33

Author(s):

Stéphane Loisel ◽

Pierrick Piette ◽

Cheng-Hsien Jason Tsai

Keyword(s):

Machine Learning ◽

Risk Management ◽

Regression Tree ◽

Classification Problem ◽

Point Of View ◽

Classification And Regression Tree ◽

Gradient Boosting ◽

Support Vector ◽

Learning Approaches ◽

Extreme Gradient Boosting

Abstract Modeling policyholders’ lapse behaviors is important to a life insurer, since lapses affect pricing, reserving, profitability, liquidity, risk management, and the solvency of the insurer. In this paper, we apply two machine learning methods to lapse modeling. Then, we evaluate the performance of these two methods along with two popular statistical methods by means of statistical accuracy and profitability measure. Moreover, we adopt an innovative point of view on the lapse prediction problem that comes from churn management. We transform the classification problem into a regression question and then perform optimization, which is new to lapse risk management. We apply the aforementioned four methods to a large real-world insurance dataset. The results show that Extreme Gradient Boosting (XGBoost) and support vector machine outperform logistic regression (LR) and classification and regression tree with respect to statistic accuracy, while LR performs as well as XGBoost in terms of retention gains. This highlights the importance of a proper validation metric when comparing different methods. The optimization after the transformation brings out significant and consistent increases in economic gains. Therefore, the insurer should conduct optimization on its economic objective to achieve optimal lapse management.

Download Full-text