A Novel Method for Calculating the Parametric Hypervolume Indicator

2021 ◽  
Author(s):  
Jonathan M. Weaver-Rosen ◽  
Richard J. Malak
Keyword(s):  
Parametric Optimization ◽  
Parametric Design ◽  
Computational Time ◽  
Support Vector ◽  
The Novel ◽  
Pareto Dominance ◽  
Multi Objective ◽  
Parametric Data ◽  
Novel Method ◽  
Hypervolume Indicator

Abstract This paper presents a new methodology for calculating the hypervolume indicator (HVI)for multi-objective and parametric data. Existing multi-objective HVI calculation techniques cannot be directly used for parametric data because designers do not have preferences for parameters like they do for objectives. The novel method presented herein allows for the consideration of both objectives and parameters through the newly introduced hypercone heuristic (HCH). This heuristic relaxes the strict rules of parametric Pareto dominance for a more practical dominance assessment when comparing designs of differing parameter values without violating Pareto dominance rules. A parametric HVI (pHVI) enhances a design engineer’s toolkit by enabling both online and offline evaluation of parametric optimization results. The pHVI measure allows designers to compare solution sets, detect optimization convergence, and to better inform optimization procedures in a parametric context. Results show that the HCH-based pHVI yields a similar quality measure to the existing technique based on a support vector domain description (SVDD) in a fraction of the computational time. Furthermore, the novel HCH-based pHVI technique satisfies multi-objective HVI properties allowing previous applications of the HVI to be applied to multi-objective parametric optimization. This contribution enables the field of parametric optimization, and thus parametric design, to benefit from prior and future advances in the multi-objective optimization domain involving the HVI.

Artificial Intelligence for the Online Prediction of the Cool-down Time in a Subsea Pipeline After an Unplanned Shutdown

2021 ◽  
Author(s):  
Alberto Gerri ◽  
Ahmed Shokry ◽  
Enrico Zio ◽  
Marco Montini
Keyword(s):  
High Risk ◽  
Operating Conditions ◽  
Low Risk ◽  
Minimum Time ◽  
Computational Time ◽  
Risk Level ◽  
Support Vector ◽  
Svm Classifier ◽  
The Novel ◽  
Subsea Pipelines

Abstract Hydrates formation in subsea pipelines is one of the main reliability concerns for flow assurance engineers. A fast and reliable assessment of the Cool-Down Time (CDT), the period between a shut-down event and possible hydrates formation in the asset, is of key importance for the safety of operations. Existing methods for the CDT prediction are highly dependent on the use of very complex physics-based models that demand large computational time, which hinders their usage in an online environment. Therefore, this work presents a novel methodology for the development of surrogate models that predict, in a fast and accurate way, the CDT in subsea pipelines after unplanned shutdowns. The proposed methodology is, innovatively, tailored on the basis of reliability perspective, by treating the CDT as a risk index, where a critic CDT threshold (i.e. the minimum time needed by the operator to preserve the line from hydrates formation) is considered to distinguish the simulation outputs into high-risk and low-risk domains. The methodology relies on the development of a hybrid Machine Learning (ML) based model using datasets generated through complex physics-based model’ simulations. The hybrid ML-based model consists of a Support Vector Machine (SVM) classifier that assigns a risk level (high or low) to the measured operating condition of the asset, and two Artificial Neural Networks (ANNs) for predicting the CDT at the high-risk (low CDT) or the low-risk (high CDT) operating conditions previously assigned by the classifier. The effectiveness of the proposed methodology is validated by its application to a case study involving a pipeline in an offshore western African asset, modelled by a transient physics-based commercial software. The results show outperformance of the capabilities of the proposed hybrid ML-based model (i.e., SVM + 2 ANNs) compared to the classical approach (i.e. modelling the entire system with one global ANN) in terms of enhancing the prediction of the CDT during the high-risk conditions of the asset. This behaviour is confirmed applying the novel methodology to training datasets of different size. In fact, the high-risk Normalized Root Mean Square Error (NRMSE) is reduced on average of 15% compared to the NRMSE of a global ANN model. Moreover, it’s shown that high-risk CDT are better predicted by the hybrid model even if the critic CDT, which divides the simulation outputs in high-risk and low-risk values (i.e. the minimum time needed by the operator to preserve the line from hydrates formation), changes. The enhancement, in this case, is on average of 14.6%. Eventually, results show how the novel methodology cuts down by more than one hundred seventy-eight times the computational times for online CDT predictions compared to the physics-based model.

An Improved Online SVM Algorithm for Overcoming the Influence of Muscle Fatigue in sEMG Based Human Machine Interaction

2014 ◽  
Vol 536-537 ◽  
pp. 1026-1031
Author(s):  
Jun Cai ◽  
Rui Liu
Keyword(s):  
Muscle Fatigue ◽  
Training Data ◽  
Support Vector ◽  
The Novel ◽  
Human Machine Interaction ◽  
Svm Algorithm ◽  
Svm Model ◽  
Novel Method ◽  
Machine Interface ◽  
The Stability

For the problem that the stability of surface Electromyographic EMG(sEMG) based human-machine interface(HMI) declines as the muscle fatigue takes place, an improved incremental training algorithm for online support vector machine(SVM) is proposed. This paper study the changes of sEMG when muscle fatigue occurs by the method of continuous wavelet transform, and then apply the improved online SVM for sEMG classification. The novel method adjusts the parameters of SVM model to adapt itself based on the changes of sEMG signals and the training data is conditionally selected and forgot. Experiment results show that the presented algorithm performs high modeling precision and training speed is increased. Furthermore, this method effectively overcomes the influence of muscle fatigue during long-term operating sEMG based HMI.

A Novel SVM-Based Method for Seismic First-Arrival Detecting

2010 ◽  
Vol 29-32 ◽  
pp. 973-978 ◽  
Author(s):  
Ming Chen ◽  
Yong Li ◽  
Jun Xie
Keyword(s):  
Support Vector ◽  
The Novel ◽  
Ann Model ◽  
Seismic Record ◽  
Feature Vectors ◽  
Svm Model ◽  
First Arrival ◽  
Ann Models ◽  
Novel Method ◽  
Artificial Neural Network Ann

First arrivals detecting on seismic record is important at all times. A novel support vector machine (SVM)-based method for seismic first-arrival pickup is proposed in this research. Firstly, the multi-resolution wavelet decomposition is used to de-noise the seismic record. And then, feature vectors are extracted from the denoise data. Finally, both SVM and artificial neural network (ANN) models are employed to train and predict the feature vectors. Experimental results demonstrate that the SVM model gives better accuracy than the ANN model. It is promising that the novel method is very prospective.

scholarly journals Anemia Selection in Pregnant Women by using Random prediction (Rp) Classification Algorithm

2019 ◽  
Vol 8 (2) ◽  
pp. 2623-2630 ◽  
Keyword(s):  
Data Mining ◽  
Feature Selection ◽  
Classification Algorithm ◽  
Computational Time ◽  
The Novel ◽  
Data Mining Algorithms ◽  
Novel Method ◽  
Mining Algorithms ◽  
Median Vector ◽  
Selection Of

Anemia is the global hematological disorder that occurs in pregnancy. The feature selection of unknown logical knowledge from the large dataset is capable with data mining techniques. The paper evaluates anemia features classes of Non-anemic, Mild and Severe or moderate in real time large-dimensional dataset. In the previous works, Anemia diseases can be classified in a selection of approaches, based on the Artificial Neural Networks (ANN), Gausnominal Classification and VectNeighbour classification. In these previous studies attains the proper feature selection with classification accuracy but it takes large time to predict the feature selection. So the current paper to overcome the feature selection, computational time process presents an improved Median vector feature selection (IMVFS) algorithm and new RandomPrediction (RP) classification algorithm to predict the anemia disease classes (Mild, Not anemic and Severe and moderate) based on the data mining algorithms. The results have shown that the performance of the novel method is effective compared with our previous Classification of ANN, Gausnominal and VectNeighbour classification algorithms. As the Experimental results show that proposed RandomPrediction (RP) classification with (IMVFS) feature selection methods clearly outperform than our previous methods

Generic Topology Optimization Based on Local State Features

2017 ◽  
Author(s):  
Nikola Aulig
Keyword(s):  
Topology Optimization ◽  
Physical Simulation ◽  
Support Vector ◽  
The Novel ◽  
Simulation Data ◽  
Innovative Solutions ◽  
Physical Simulations ◽  
Product Design Process ◽  
Novel Method ◽  
Vehicle Crashworthiness

The work at hand addresses engineers, designers and scientists who face the challenging Task of devising concept structures in a virtual product design process that involves more and more sophisticated physical simulations. Using methods of evolutionary optimization and machine learning, this dissertation explores a novel generic topology optimization algorithm, which is able to provide concept designs even for problems involving complex, black-box simulations. A self-contained learning component utilizes physical simulation data to generate a search direction. The generic topology optimization is studied in conjunction with statistical models such as neural networks or support vector regression. In empirical experiments, the novel method reproduces reference structures with Minimum compliance and provides innovative solutions in the domain of vehicle crashworthiness optimization. Contents Symbols and Abbreviations XIV Abstract XVII Zusammenfassung XVIII 1 Introduction 1 2 Fundamentals...

Research on the Spatial Error Distribution Model of Flexible Coordinate Measuring Machines Based on Support Vector Regression

2011 ◽  
Vol 201-203 ◽  
pp. 2058-2062
Author(s):  
Da Teng Zheng ◽  
Ye Tai Fei ◽  
Yong Gang Hu ◽  
Rui Chang Yang
Keyword(s):  
Support Vector Regression ◽  
Measurement Accuracy ◽  
Error Distribution ◽  
Distribution Model ◽  
Support Vector ◽  
The Novel ◽  
Coordinate Measuring Machines ◽  
Spatial Error ◽  
Novel Method ◽  
Measurement Area

One of the main reasons on Flexible Coordinate Measuring Machines’ (FCMM) less application is that its accuracy can’t meet the practical requirements. The spatial error distribution model of FCMM is build using Support Vector Regression (SVR) method. The test shows the data from the model can effectively fit the samples, and arguably the model based on SVR is feasible. The novel method gives an idea of studying the spatial error distribution model of FCMM. Futhermore, the model can provide reasearch foundations for definiting optimal measurement area and thus help to improve the measurement accuracy in the next work.

Take full advantage of unlabeled data for sentiment classification

Kybernetes ◽  
2018 ◽  
Vol 47 (3) ◽  
pp. 474-486 ◽  
Author(s):  
Lei La ◽  
Shuyan Cao ◽  
Liangjuan Qin
Keyword(s):  
Social Network ◽  
Supervised Learning ◽  
Sentiment Classification ◽  
Unlabeled Data ◽  
Computational Time ◽  
The Novel ◽  
Related Field ◽  
Content Type ◽  
Sentiment Mining ◽  
Novel Method

Purpose As a foundational issue of social mining, sentiment classification suffered from a lack of unlabeled data. To enhance accuracy of classification with few labeled data, many semi-supervised algorithms had been proposed. These algorithms improved the classification performance when the labeled data are insufficient. However, precision and efficiency are difficult to be ensured at the same time in many semi-supervised methods. This paper aims to present a novel method for using unlabeled data in a more accurate and more efficient way. Design/methodology/approach First, the authors designed a boosting-based method for unlabeled data selection. The improved boosting-based method can choose unlabeled data which have the same distribution with the labeled data. The authors then proposed a novel strategy which can combine weak classifiers into strong classifiers that are more rational. Finally, a semi-supervised sentiment classification algorithm is given. Findings Experimental results demonstrate that the novel algorithm can achieve really high accuracy with low time consumption. It is helpful for achieving high-performance social network-related applications. Research limitations/implications The novel method needs a small labeled data set for semi-supervised learning. Maybe someday the authors can improve it to an unsupervised method. Practical implications The mentioned method can be used in text mining, image classification, audio processing and so on, and also in an unstructured data mining-related field. Overcome the problem of insufficient labeled data and achieve high precision using fewer computational time. Social implications Sentiment mining has wide applications in public opinion management, public security, market analysis, social network and related fields. Sentiment classification is the basis of sentiment mining. Originality/value According to what the authors have been informed, it is the first time transfer learning be introduced to AdaBoost for semi-supervised learning. Moreover, the improved AdaBoost uses a totally new mechanism for weighting.

scholarly journals A Hybrid Approach for Fault Diagnosis of Railway Rolling Bearings Using STWD-EMD-GA-LSSVM

2016 ◽  
Vol 2016 ◽  
pp. 1-7 ◽  
Author(s):  
Dechen Yao ◽  
Jianwei Yang ◽  
Xi Li ◽  
Chunqing Zhao
Keyword(s):  
Fault Diagnosis ◽  
Hybrid Approach ◽  
Rolling Bearing ◽  
Support Vector ◽  
The Novel ◽  
Intrinsic Mode Functions ◽  
Rolling Bearings ◽  
Vibration Signals ◽  
Mode Decomposition ◽  
Novel Method

Vibration signals resulting from railway rolling bearings are nonstationary by nature; this paper proposes a hybrid approach for the fault diagnosis of railway rolling bearings using segment threshold wavelet denoising (STWD), empirical mode decomposition (EMD), genetic algorithm (GA), and least squares support vector machine (LSSVM). The original signal is first denoised using STWD as a prefilter, which improves the subsequent decomposition into a number of intrinsic mode functions (IMFs) using EMD. Secondly, the IMF energy-torques are extracted as feature parameters. Concurrently, a GA is employed to optimize the LSSVM to improve the classification accuracy. Finally, the extracted features are used as inputs for classification by the GA-LSSVM. Actual railway rolling bearing vibration signals are used to experimentally verify the effectiveness of the proposed method. The results show that the novel method is effective and accurate for fault diagnosis of railway rolling bearings.

A novel method for the determination of linting propensity of paper

TAPPI Journal ◽  
2012 ◽  
Vol 11 (10) ◽  
pp. 9-17
Author(s):  
ALESSANDRA GERLI ◽  
LEENDERT C. EIGENBROOD
Keyword(s):  
Ash Content ◽  
The Novel ◽  
Offset Printing ◽  
Excellent Correlation ◽  
End Conditions ◽  
Bottom Side ◽  
Novel Method ◽  
Total Fraction

A novel method was developed for the determination of linting propensity of paper based on printing with an IGT printability tester and image analysis of the printed strips. On average, the total fraction of the surface removed as lint during printing is 0.01%-0.1%. This value is lower than those reported in most laboratory printing tests, and more representative of commercial offset printing applications. Newsprint paper produced on a roll/blade former machine was evaluated for linting propensity using the novel method and also printed on a commercial coldset offset press. Laboratory and commercial printing results matched well, showing that linting was higher for the bottom side of paper than for the top side, and that linting could be reduced on both sides by application of a dry-strength additive. In a second case study, varying wet-end conditions were used on a hybrid former machine to produce four paper reels, with the goal of matching the low linting propensity of the paper produced on a machine with gap former configuration. We found that the retention program, by improving fiber fines retention, substantially reduced the linting propensity of the paper produced on the hybrid former machine. The papers were also printed on a commercial coldset offset press. An excellent correlation was found between the total lint area removed from the bottom side of the paper samples during laboratory printing and lint collected on halftone areas of the first upper printing unit after 45000 copies. Finally, the method was applied to determine the linting propensity of highly filled supercalendered paper produced on a hybrid former machine. In this case, the linting propensity of the bottom side of paper correlated with its ash content.

Sign in / Sign up

Export Citation Format

Share Document