Machine learning with applications最新文献

{"title":"Application of machine learning for seam profile identification in robotic welding","authors":"Fatemeh Habibkhah,&nbsp;Mehrdad Moallem","doi":"10.1016/j.mlwa.2025.100633","DOIUrl":"10.1016/j.mlwa.2025.100633","url":null,"abstract":"<div><div>This paper addresses critical challenges in automated robotic welding, emphasizing precise weld groove profiling for pipe welding applications. By integrating advanced laser scanning technology with the Local Outlier Factor (LOF) algorithm, the research effectively mitigates outliers and compensates for incomplete data—persistent issues in dynamic manufacturing environments. To further enhance accuracy, a robust neural network model is employed to predict weld groove alignment, a crucial factor in maintaining weld structural integrity. The LOF algorithm was chosen for its ability to detect spatial anomalies, ensuring the exclusion of erroneous data that could compromise welding precision. Experimental results demonstrate that the combined use of LOF and neural networks significantly improves the operational efficiency of robotic welding, delivering consistently strong and precise welds across diverse manufacturing scenarios. The model achieved an average mean square error of 0.078 and an R² value of 0.995, accurately predicting 99.5 % of data. Therefore, neural network modeling enables accurate interpolation of missing data and real-time adjustments to varying operational conditions.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"20 ","pages":"Article 100633"},"PeriodicalIF":0.0,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143512090","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Uncertainty quantification based on symbolic regression and probabilistic programming and its application","authors":"Yuyang Zhao ,&nbsp;Hongbo Zhao","doi":"10.1016/j.mlwa.2025.100632","DOIUrl":"10.1016/j.mlwa.2025.100632","url":null,"abstract":"<div><div>The joint roughness coefficient (JRC) is critical to evaluate the strength and deformation behavior of joint rock mass in rock engineering. Various methods have been developed to estimate JRC value based on the statistical parameter of rock joints. The JRC value is uncertain due to the complex, random rock joint. Uncertainty is an essential characteristic of rock joints. However, the traditional determinative method cannot deal with uncertainty during the analysis, evaluation, and characterization of the mechanism for the rock joint. This study developed a novel JRC determination framework to estimate the JRC value and evaluate the uncertainty of rock joints based on symbolic regression and probabilistic programming. The symbolic regression was utilized to generate the general empirical equation with the unknown coefficient for the JRC determination of rock joints. The probabilistic programming was used to quantify the uncertainty of the rock joint roughness. The ten standard rock joint profiles illustrated and investigated the developed framework. And then, the developed framework was applied to the collected rock joint profile from the literature. The predicted JRC value was compared with the traditional empirical equations. The results show that the generalization performance of the developed framework is better than the traditional determinative empirical equation. It provides a scientific, reliable, and helpful to estimate the JRC value and characterize the mechanical behavior of joint rock mass.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"20 ","pages":"Article 100632"},"PeriodicalIF":0.0,"publicationDate":"2025-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143519591","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predicting classification errors using NLP-based machine learning algorithms and expert opinions","authors":"Peiheng Gao ,&nbsp;Chen Yang ,&nbsp;Ning Sun ,&nbsp;Ričardas Zitikis","doi":"10.1016/j.mlwa.2025.100630","DOIUrl":"10.1016/j.mlwa.2025.100630","url":null,"abstract":"<div><div>Various intentional and unintentional biases of humans manifest in classification tasks, such as those related to risk management. In this paper we demonstrate the role of ML algorithms when accomplishing these tasks and highlight the role of expert know-how when training the staff as well as, and very importantly, when training and fine-tuning ML algorithms. In the process of doing so and when facing well-known inefficiencies of the traditional F1 score, especially when working with unbalanced datasets, we suggest a modification of the score by incorporating human-experience-trained algorithms, which include both expert-trained algorithms (i.e., with the involvement of expert experiences in classification tasks) and staff-trained algorithms (i.e., with the involvement of experiences of those staff who have been trained by experts). Our findings reveal that the modified F1 score diverges from the traditional staff F1 score when the staff labels exhibit weak correlation with expert labels, which indicates insufficient staff training. Furthermore, the Long Short-Term Memory (LSTM) model outperforms other classifiers in terms of the modified F1 score when applied to the classification of textual narratives in consumer complaints.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"19 ","pages":"Article 100630"},"PeriodicalIF":0.0,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143436468","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning-based state of charge estimation: A comparison between CatBoost model and C-BLSTM-AE model","authors":"Abderrahim Zilali ,&nbsp;Mehdi Adda ,&nbsp;Khaled Ziane ,&nbsp;Maxime Berger","doi":"10.1016/j.mlwa.2025.100629","DOIUrl":"10.1016/j.mlwa.2025.100629","url":null,"abstract":"<div><div>The State of Charge (SOC) is a key metric within a Lithium-ion battery management system (BMS). Accurate SOC estimation is essential for enhancing battery longevity and ensuring user safety, making it a critical component of an effective BMS. Although SOC estimation has become an active research area for the machine learning (ML) community, only a handful of works have considered its estimation at negative temperatures. This paper proposes the application of two machine learning-based approaches for SOC estimation that perform well at wide range of temperatures (positive and negative) and varying dynamic loads. The first one is a hybrid deep learning approach based on the Convolutional BLSTM Auto-Encoder (C-BLSTM-AE) model that relies on extracting abstract features from input data. The second one is a CatBoost model that leverages the gradient boosting technique to enhance the prediction made by its constituent trees. The performance of the models is evaluated by comparing their regression accuracy and computational resource utilization. The C-BLSTM-AE model achieves a low Mean Absolute Error (MAE) of <strong>0.52 %</strong> under fixed ambient temperature conditions and maintains a MAE of <strong>1.03 %</strong> for variable ambient temperatures. The CatBoost model achieves a MAE of <strong>0.69 %</strong> with fixed temperature settings and a MAE of <strong>1.09 %</strong> under variable temperature conditions.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"20 ","pages":"Article 100629"},"PeriodicalIF":0.0,"publicationDate":"2025-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143519593","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A deep learning framework for accurate COVID-19 classification in CT-scan images","authors":"Shirin Kordnoori ,&nbsp;Maliheh Sabeti ,&nbsp;Hamidreza Mostafaei ,&nbsp;Saeed Seyed Agha Banihashemi","doi":"10.1016/j.mlwa.2025.100628","DOIUrl":"10.1016/j.mlwa.2025.100628","url":null,"abstract":"<div><h3>Background</h3><div>In response to the global COVID-19 pandemic, we have introduced a binary classification model that employs convolutional layers to differentiate between normal cases and COVID-19-infected cases. Our primary aim was to address the urgent need for a highly efficient and accurate diagnostic tool to combat the widespread outbreak of COVID-19.</div></div><div><h3>Methods</h3><div>To achieve the background, we proposed a convolutional structure that comprises 10 layers in the encoder and 3 dense layers in the decoder. We conducted comprehensive experiments and evaluations using four distinct datasets.</div></div><div><h3>Results</h3><div>The outcomes of our study consistently demonstrated remarkable performance, with our proposed model achieving an accuracy of 89.00 %, a sensitivity of 0.95, a specificity of 0.88, and an impressive AUC of 0.92. Notably, Dataset 4 yielded the most promising results among all datasets, underscoring the effectiveness of our approach.</div></div><div><h3>Conclusion</h3><div>Our research substantiates the superiority of our model over previous methodologies and pre-trained models. Furthermore, it significantly contributes to global efforts in combating COVID-19 by providing an advanced diagnostic tool. This work also paves the way for future breakthroughs in the field of medical image analysis.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"19 ","pages":"Article 100628"},"PeriodicalIF":0.0,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143419437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Noninvasive estimation of blood glucose and HbA1c using Quantum Machine Learning technique","authors":"Parama Sridevi ,&nbsp;Masud Rabbani ,&nbsp;Md Hasanul Aziz ,&nbsp;Paramita Basak Upama ,&nbsp;Sayed Mashroor Mamun ,&nbsp;Rumi Ahmed Khan ,&nbsp;Sheikh Iqbal Ahamed","doi":"10.1016/j.mlwa.2025.100626","DOIUrl":"10.1016/j.mlwa.2025.100626","url":null,"abstract":"<div><div>In this paper, we developed models with quantum and classical machine learning algorithms to detect blood glucose and HbA1c noninvasively from ten-second fingertip video by deploying a smartphone and near-infrared spectroscopy. Using our developed framework, we collected 136 participants’ ten-second fingertip videos with their baseline blood glucose and HbA1c levels after getting approval from the Institutional Review Board (IRB). We extracted 45 PPG (photoplethysmography) features from the ten-second fingertip video by using the Beer–Lambert law and applied feature engineering to select the most important features. We applied two Quantum Machine Learning (QML) based algorithms and seven Classical Machine Learning (CML) based algorithms for estimating blood glucose and HbA1c levels. The application of QML for the noninvasive estimation of blood glucose and HbA1c is a new and unexplored research area. Among all developed models, the Quantum Support Vector Machine performs best for predicting both blood glucose and HbA1c. The Quantum Support Vector Machine provides an accuracy of 89.30% and an average k-fold cross-validation score of 92.50% for blood glucose prediction and an accuracy of 96.30% and an average k-fold cross-validation score of 92.50% for HbA1c prediction. Our study signifies the potential of QML algorithms in noninvasive health monitoring, especially in the less-explored area of blood glucose and HbA1c estimation. The high performance of the developed models paves the way for advancing noninvasive techniques for measuring blood constituents. These findings offer promising applications in personalized healthcare, including continuous monitoring, early disease diagnosis, and more convenient management of chronic conditions.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"19 ","pages":"Article 100626"},"PeriodicalIF":0.0,"publicationDate":"2025-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143386451","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multiobjective continuation method to compute the regularization path of deep neural networks","authors":"Augustina Chidinma Amakor ,&nbsp;Konstantin Sonntag ,&nbsp;Sebastian Peitz","doi":"10.1016/j.mlwa.2025.100625","DOIUrl":"10.1016/j.mlwa.2025.100625","url":null,"abstract":"<div><div>Sparsity is a highly desired feature in deep neural networks (DNNs) since it ensures numerical efficiency, improves the interpretability (due to the smaller number of relevant features), and robustness. For linear models, it is well known that there exists a <em>regularization path</em> connecting the sparsest solution in terms of the <span><math><msup><mrow><mi>ℓ</mi></mrow><mrow><mn>1</mn></mrow></msup></math></span> norm, i.e., zero weights and the non-regularized solution. Recently, there was a first attempt to extend the concept of regularization paths to DNNs by means of treating the empirical loss and sparsity (<span><math><msup><mrow><mi>ℓ</mi></mrow><mrow><mn>1</mn></mrow></msup></math></span> norm) as two conflicting criteria and solving the resulting multiobjective optimization problem. However, due to the non-smoothness of the <span><math><msup><mrow><mi>ℓ</mi></mrow><mrow><mn>1</mn></mrow></msup></math></span> norm and the large number of parameters, this approach is not very efficient from a computational perspective. To overcome this limitation, we present an algorithm that allows for the approximation of the entire Pareto front for the above-mentioned objectives in a very efficient manner for high-dimensional DNNs with millions of parameters. We present numerical examples using both deterministic and stochastic gradients. We furthermore demonstrate that knowledge of the regularization path allows for a well-generalizing network parametrization.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"19 ","pages":"Article 100625"},"PeriodicalIF":0.0,"publicationDate":"2025-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143170236","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Apply a deep learning hybrid model optimized by an Improved Chimp Optimization Algorithm in PM2.5 prediction","authors":"Ming Wei,&nbsp;Xiaopeng Du","doi":"10.1016/j.mlwa.2025.100624","DOIUrl":"10.1016/j.mlwa.2025.100624","url":null,"abstract":"<div><div>PM_2.5 pollution in the atmosphere not only contaminates the environment but also seriously affects human health. Therefore, studying how to accurately predict future PM_2.5 concentrations holds significant importance and practical value. This paper innovatively <span><math><mrow><mi>P</mi><msub><mrow><mi>M</mi></mrow><mrow><mn>2</mn><mo>.</mo><mn>5</mn></mrow></msub></mrow></math></span>proposes a high-accuracy prediction model: RF-ICHOA-CNN-LSTM-Attention. First, the Random Forest (RF) model is utilized to evaluate the importance of air pollution and meteorological features and select more suitable input features. Subsequently, a one-dimensional convolutional neural network (1DCNN) with efficient feature extraction capability is used to extract dynamic features from sequences. The extracted feature vector sequences are then fed into a Long Short-Term Memory Network (LSTM). After the LSTM, an Attention Mechanism is incorporated to assign different weights to the input features, emphasizing the role of the important features. Additionally, the Improved Chimp Optimization Algorithm (IChOA) is employed to optimize the number of neurons in the two hidden layers of LSTM, the learning rate, and the number of training epochs. The experimental results on 12 test functions demonstrate that the optimization performance of IChOA is better than that of ChOA and the representative swarm optimization algorithms used for comparison. In the case of PM_2.5 predictions in Yining and Beijing, experimental results show that the proposed model achieved the best performance in terms of RMSE, MAE, and R<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> This indicates its excellent prediction accuracy and generalization capability, Thus proving its effectiveness in predicting PM_2.5 concentration in the real world.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"19 ","pages":"Article 100624"},"PeriodicalIF":0.0,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143170235","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ensembles of deep one-class classifiers for multi-class image classification","authors":"Alexander Novotny ,&nbsp;George Bebis ,&nbsp;Alireza Tavakkoli ,&nbsp;Mircea Nicolescu","doi":"10.1016/j.mlwa.2025.100621","DOIUrl":"10.1016/j.mlwa.2025.100621","url":null,"abstract":"<div><div>Traditional methods for multi-class classification (MCC) involve using a monolithic feature extractor and classifier trained on data from all the classes simultaneously. These methods are dependent on the number and types of classes and are therefore rigid against changes to the class structure. For instance, if the number of classes needs to be modified or new training data becomes available, retraining would be required for optimum classification performance. Moreover, these classifiers can become biased toward classes with a large data imbalance. An alternative, more attractive framework is to consider an ensemble of one-class classifiers (EOCC) where each one-class classifier (OCC) is trained with data from a single class only, without using any information from the other classes. Although this framework has not yet systematically matched or surpassed the performance of traditional MCC approaches, it deserves further investigation for several reasons. First, it provides a more flexible framework for handling changes in class structure compared to the traditional MCC approach. Second, it is less biased toward classes with large data imbalances compared to the multi-class classification approach. Finally, each OCC can be separately optimized depending on the characteristics of the class it represents. In this paper, we have performed extensive experiments to evaluate EOCC for MCC using traditional OCCs based on Principal Component Analysis (PCA) and Auto-encoders (AE) as well as newly proposed OCCs based on Generative Adversarial Networks (GANs). Moreover, we have compared the performance of EOCC with traditional multi-class DL classifiers including VGG-19, Resnet and EfficientNet. Two different datasets were used in our experiments: (i) a subset from the Plant Village dataset plant disease dataset with high variance in the number of classes and amount of data in each class, and (ii) an Alzheimer’s disease dataset with low amounts of data and a large imbalance in data between classes. Our results show that the GAN-based EOCC outperform previous EOCC approaches and improve the performance gap with traditional MCC approaches.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"19 ","pages":"Article 100621"},"PeriodicalIF":0.0,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143170233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Safety analysis in the era of large language models: A case study of STPA using ChatGPT","authors":"Yi Qi ,&nbsp;Xingyu Zhao ,&nbsp;Siddartha Khastgir ,&nbsp;Xiaowei Huang","doi":"10.1016/j.mlwa.2025.100622","DOIUrl":"10.1016/j.mlwa.2025.100622","url":null,"abstract":"<div><div>Can safety analysis leverage Large Language Models (LLMs)? This study examines the application of Systems Theoretic Process Analysis (STPA) to Automatic Emergency Brake (AEB) and Electricity Demand Side Management (DSM) systems, utilising Chat Generative Pre-Trained Transformer (ChatGPT). We investigate the impact of collaboration schemes, input semantic complexity, and prompt engineering on STPA results. Comparative results indicate that using ChatGPT without human intervention may be inadequate due to reliability issues. However, with careful design, it has the potential to outperform human experts. No statistically significant differences were observed when varying the input semantic complexity or using domain-agnostic prompt guidelines. While STPA-specific prompt engineering produced statistically significant and more pertinent results, ChatGPT generally yielded more conservative and less comprehensive outcomes. We also identify future challenges, such as concerns regarding the trustworthiness of LLMs and the need for standardisation and regulation in this field. All experimental data are publicly accessible.</div></div>","PeriodicalId":74093,"journal":{"name":"Machine learning with applications","volume":"19 ","pages":"Article 100622"},"PeriodicalIF":0.0,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143170234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}