Adaptive learning for reliability analysis using Support Vector Machines

Pepper, Nick; Crespo, Luis G.; Montomoli, Francesco

doi:10.1016/j.ress.2022.108635

Cited by 28 publications

(5 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The SVM belongs to a family of generalised linear classifiers, such as backpropagation neural network, RBF, learning vector quantisation and Kohonen. This family of classifiers can both minimise the empirical classification error and maximise the geometric margin 34 . The SVM employs structural risk minimisation to find the best hyperplane that separates two different classes in the input space of the SVM.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Combination of metal oxide semiconductor gas sensor array and solid‐phase microextraction gas chromatography–mass spectrometry for odour classification of brewed coffee

Hardoyono,

Windhani

2023

Flavour & Fragrance J

View full text Add to dashboard Cite

Odour analysis of coffee using low‐cost and portable instruments in coffee shops, restaurants and bars is essential to keep the loyalty of coffee consumers. This paper aimed to analyse the performance of a gas sensor array for odour classification of brewed coffee. Five grams of ground coffee sample from five different brands was brewed in 80 mL of hot water at a temperature of 90°C. The gas sensor array then measured the sensor's response to the brewed coffee odour. The recorded data were analysed using a principal component analysis (PCA), a hierarchical cluster analysis (HCA) and a support vector machine (SVM). Solid‐phase microextraction gas chromatography–mass spectrometry (SPME‐GC‐MS) was used to identify the five coffee samples' volatile organic compounds (VOCs). The visualisation of the PCA score plot shows that the gas sensor array efficiently classifies the brewed coffee based on different odours. The SVM classification using a polynomial kernel obtained an accuracy of 95.21% using training data sets and an accuracy of 96.94% using testing data sets. Meanwhile, for SVM classification using radial basis function kernel, the SVM obtained an accuracy of 100% for training data sets and 93.06% for testing data sets. The SPME‐GC‐MS analysis showed that the abundance of 2‐furanmethanol; 2‐methoxy‐4‐vinyl phenol; phenol, 4‐ethyl‐2‐methoxy‐ and acetic acid contributed to the separation of the first and the second clusters in the principal components coordinate. Based on data analysis, the gas sensor showed high performance as a low‐cost and portable instrument for odour analysis of coffee based on sensory technique.

show abstract

Section: Resultsmentioning

confidence: 99%

“…This family of classifiers can both minimise the empirical classification error and maximise the geometric margin. 34 The SVM employs structural risk minimisation to find the best hyperplane that separates two different classes in the input space of the SVM. This algorithm obtains high accuracy in robust classification.…”

Section: Svm Analysismentioning

confidence: 99%

Combination of metal oxide semiconductor gas sensor array and solid‐phase microextraction gas chromatography–mass spectrometry for odour classification of brewed coffee

Hardoyono,

Windhani

2023

Flavour & Fragrance J

View full text Add to dashboard Cite

show abstract

“…The learning rate is maintained for each network weight (parameter) and is adapted separately as learning progresses. The method calculates the individual adaptive learning rates (Pepper et al, 2022), (Kotsyuba et al, 2022) for different parameters of the first and second moment estimates of the gradient.…”

Section: Literature Reviewmentioning

confidence: 99%

Disease Classification on Rice Leaves using DenseNet121, DenseNet169, DenseNet201

Saputra

Hindarto

Santoso

2023

SinkrOn

View full text Add to dashboard Cite

Rice is a plant that can grow in the tropics. This plant can produce food that can meet the needs of the people of a country. This plant can grow well if it is cared for properly. If the planting has used good care, such as providing adequate water, adding good fertilizer, it can be ascertained that it will produce a lot of rice fruit after harvesting. This often causes concern if rice growers have given good care but often produce less rice fruit because rice plants are attacked by various diseases. This is what makes the problem, that rice plants are attacked by diseases. Before spraying diseases or pests, farmers should have an understanding of diseases in rice. This makes farmers not wrong in choosing drugs for farmers' rice. It is very vulnerable if farmers do not know about the rice disease. Therefore, it is necessary to observe what types of rice diseases attack rice plants. Observations are not enough just to take pictures with a camera. But it is necessary to carry out further analysis of rice diseases. The presence of information technology is now able to recognize any type. One of the machine learning technologies is able to detect rice diseases. One of these branches of machine learning is deep learning. By using a dataset that focuses on rice disease, the model generated from deep learning training is able to detect rice disease. The purpose of this research is to predict disease in rice leaves using deep learning, namely DenseNet. Training using DenseNet, namely DenseNet121, DenseNet169 and DenseNet201. Accuracy using DenseNet121 reached 91.67%, DenseNet169 reached 90%, and DenseNet201 reached 88.33%. The model training time takes 24 seconds.

show abstract

“…Surrogate model 26 is an effective method to improve the computational efficiency of MCS, and its accuracy and efficiency depend on the model type and sampling strategy. Commonly used methods include the Kriging function, [6][7][8] support vector machine, 27,28 and neural network. 29,30 There is no clear conclusion on which method is better than others for all problems.…”

Section: Introductionmentioning

confidence: 99%

Action reliability analysis under random‐interval hybrid uncertainty of chain conveyor based on adaptive intelligent extremum surrogate model

Wen,

Hou,

Lin

et al. 2023

Quality & Reliability Eng

View full text Add to dashboard Cite

The action responses of controlled mechanisms are often multiparametric, nonlinear, and uncertain. Complex dynamics and limited uncertain information pose difficulties for action reliability analysis. This paper develops an adaptive intelligent extremum surrogate model (AIESM) method for the action reliability under random‐interval hybrid uncertainty of a chain conveyor. First, a dynamic model of the chain conveyor is established, which considers the impact and frictional effects within the mechanical system, the regulating effects of the control system, and the external disturbances of the system. After that, a hybrid kernel extreme learning machine optimized by the sparrow search algorithm is employed as an intelligent surrogate model to construct the initial surrogate model from the hybrid uncertain variables to the limit state function (LSF) response and the extremum surrogate model (ESM) from the random variables to the LSF extremum response. An adaptive infilling strategy combining active learning and opposition‐based learning is applied to improve the accuracy and efficiency of the ESM and reduce the estimation error of action reliability. Finally, the action reliability interval bounds are obtained by Monte Carlo simulation based on the ESM. Two numerical examples are presented to illustrate the validity of the AIESM method. The action reliability interval of the chain conveyor provided by the proposed method is [0.9706, 0.9923].

show abstract

Adaptive learning for reliability analysis using Support Vector Machines

Cited by 28 publications

References 47 publications

Combination of metal oxide semiconductor gas sensor array and solid‐phase microextraction gas chromatography–mass spectrometry for odour classification of brewed coffee

Combination of metal oxide semiconductor gas sensor array and solid‐phase microextraction gas chromatography–mass spectrometry for odour classification of brewed coffee

Disease Classification on Rice Leaves using DenseNet121, DenseNet169, DenseNet201

Action reliability analysis under random‐interval hybrid uncertainty of chain conveyor based on adaptive intelligent extremum surrogate model

Contact Info

Product

Resources

About