A Review of Modern Machine Learning Techniques in the Prediction of Remaining Useful Life of Lithium-Ion Batteries

Batteries, Journal Year: 2022, Volume and Issue: 9(1), P. 13 - 13

Published: Dec. 25, 2022

The intense increase in air pollution caused by vehicular emissions is one of the main causes changing weather patterns and deteriorating health conditions. Furthermore, renewable energy sources, such as solar, wind, biofuels, suffer from supply chain-related uncertainties. electric vehicles’ powered energy, stored a battery, offers an attractive option to overcome uncertainties certain extent. development implementation cutting-edge vehicles (EVs) with long driving ranges, safety, higher reliability have been identified critical decarbonizing transportation sector. Nonetheless, capacity time usage, environmental degradation factors, end-of-life repurposing pose significant challenges usage lithium-ion batteries. In this aspect, determining battery’s remaining usable life (RUL) establishes its efficacy. It also aids testing various EV upgrades identifying factors that will improve their efficiency. Several nonlinear complicated parameters are involved process. Machine learning (ML) methodologies proven be promising tool for optimizing modeling engineering domain (non-linearity complexity). contrast scalability temporal limits battery degeneration, ML techniques provide non-invasive solution excellent accuracy minimal processing. Based on recent research, study presents objective comprehensive evaluation these challenges. RUL estimations explained detail, including examples approach applicability. many thoroughly individually studied. Finally, application-focused overview offered, emphasizing advantages terms efficiency accuracy.

Language: Английский

---

Underwater Target Detection Based on Improved YOLOv7

Journal of Marine Science and Engineering, Journal Year: 2023, Volume and Issue: 11(3), P. 677 - 677

Published: March 22, 2023

Underwater target detection is a crucial aspect of ocean exploration. However, conventional underwater methods face several challenges such as inaccurate feature extraction, slow speed, and lack robustness in complex environments. To address these limitations, this study proposes an improved YOLOv7 network (YOLOv7-AC) for detection. The proposed utilizes ACmixBlock module to replace the 3 × convolution block E-ELAN structure, incorporates jump connections 1 architecture between modules improve extraction reasoning speed. Additionally, ResNet-ACmix designed avoid information loss reduce computation, while Global Attention Mechanism (GAM) inserted backbone head parts model extraction. Furthermore, K-means++ algorithm used instead K-means obtain anchor boxes enhance accuracy. Experimental results show that outperforms original other popular methods. achieved mean average precision (mAP) value 89.6% 97.4% on URPC dataset Brackish dataset, respectively, demonstrated higher frame per second (FPS) compared model. In conclusion, represents promising solution holds great potential practical applications various tasks.

Language: Английский

---

A systematic review and meta-analysis of artificial neural network, machine learning, deep learning, and ensemble learning approaches in field of geotechnical engineering

Neural Computing and Applications, Journal Year: 2024, Volume and Issue: 36(21), P. 12655 - 12699

Published: May 13, 2024

Abstract Artificial neural networks (ANN), machine learning (ML), deep (DL), and ensemble (EL) are four outstanding approaches that enable algorithms to extract information from data make predictions or decisions autonomously without the need for direct instructions. ANN, ML, DL, EL models have found extensive application in predicting geotechnical geoenvironmental parameters. This research aims provide a comprehensive assessment of applications addressing forecasting within field related engineering, including soil mechanics, foundation rock environmental geotechnics, transportation geotechnics. Previous studies not collectively examined all algorithms—ANN, EL—and explored their advantages disadvantages engineering. categorize address this gap existing literature systematically. An dataset relevant was gathered Web Science subjected an analysis based on approach, primary focus objectives, year publication, geographical distribution, results. Additionally, study included co-occurrence keyword covered techniques, systematic reviews, review articles data, sourced Scopus database through Elsevier Journal, were then visualized using VOS Viewer further examination. The results demonstrated ANN is widely utilized despite proven potential methods engineering due real-world laboratory civil engineers often encounter. However, when it comes behavior scenarios, techniques outperform three other methods. discussed here assist understanding benefits geo area. enables practitioners select most suitable creating certainty resilient ecosystem.

Language: Английский

---

Supervised machine learning in drug discovery and development: Algorithms, applications, challenges, and prospects

Machine Learning with Applications, Journal Year: 2024, Volume and Issue: 17, P. 100576 - 100576

Published: July 24, 2024

Language: Английский

---

An Interpretable Machine Learning Approach for Hepatitis B Diagnosis

Applied Sciences, Journal Year: 2022, Volume and Issue: 12(21), P. 11127 - 11127

Published: Nov. 2, 2022

Hepatitis B is a potentially deadly liver infection caused by the hepatitis virus. It serious public health problem globally. Substantial efforts have been made to apply machine learning in detecting However, application of model interpretability limited existing literature. Model makes it easier for humans understand and trust machine-learning model. Therefore, this study, we used SHapley Additive exPlanations (SHAP), game-based theoretical approach explain visualize predictions models applied diagnosis. The algorithms building include decision tree, logistic regression, support vector machines, random forest, adaptive boosting (AdaBoost), extreme gradient (XGBoost), they achieved balanced accuracies 75%, 82%, 86%, 92%, 90%, respectively. Meanwhile, SHAP values showed that bilirubin most significant feature contributing higher mortality rate. Consequently, older patients are more likely die with elevated levels. outcome study can aid practitioners policymakers explaining result health-related problems.

Language: Английский

---

COVID-19 health data analysis and personal data preserving: A homomorphic privacy enforcement approach

Computer Communications, Journal Year: 2022, Volume and Issue: 199, P. 87 - 97

Published: Dec. 14, 2022

Language: Английский

---

An Efficient AdaBoost Algorithm with the Multiple Thresholds Classification

Applied Sciences, Journal Year: 2022, Volume and Issue: 12(12), P. 5872 - 5872

Published: June 9, 2022

Adaptive boost (AdaBoost) is a prominent example of an ensemble learning algorithm that combines weak classifiers into strong through weighted majority voting rules. AdaBoost’s classifier, with threshold classification, tries to find the best in one data dimensions, dividing two categories-1 and 1. However, some cases, this Weak Learning not accurate enough, showing poor generalization performance tendency over-fit. To solve these challenges, we first propose new classifies examples based on multiple thresholds, rather than only one, improve its accuracy. Second, paper, make changes weight allocation scheme AdaBoost use potential values other dimensions classification process, while theoretical identification provided show generality. Finally, comparative experiments between algorithms 18 datasets UCI our improved has better effect test set during training iteration.

Language: Английский

---

An ensemble learning approach for diabetes prediction using boosting techniques

Frontiers in Genetics, Journal Year: 2023, Volume and Issue: 14

Published: Oct. 26, 2023

Introduction: Diabetes is considered one of the leading healthcare concerns affecting millions worldwide. Taking appropriate action at earliest stages disease depends on early diabetes prediction and identification. To support providers for better diagnosis prognosis diseases, machine learning has been explored in industry recent years. Methods: predict diabetes, this research conducted experiments five boosting algorithms Pima dataset. The dataset was obtained from University California, Irvine (UCI) repository, which contains several important clinical features. Exploratory data analysis used to identify characteristics Moreover, upsampling, normalisation, feature selection, hyperparameter tuning were employed predictive analytics. Results: results analysed using various statistical/machine metrics k-fold cross-validation techniques. Gradient achieved greatest accuracy rate 92.85% among all classifiers. Precision, recall, f1-score, receiver operating characteristic (ROC) curves further validate model. Discussion: suggested model outperformed current studies terms accuracy, demonstrating its applicability other diseases with similar predicate indications.

Language: Английский

---

Interpretable optimisation-based approach for hyper-box classification

Machine Learning, Journal Year: 2025, Volume and Issue: 114(3)

Published: Feb. 6, 2025

Abstract Data classification is considered a fundamental research subject within the machine learning community. Researchers seek improvement of algorithms in not only accuracy, but also interpretability. Interpretable allow humans to easily understand decisions that model makes, which challenging for black box models. Mathematical programming-based have attracted considerable attention due their ability effectively compete with leading-edge terms both accuracy and Meanwhile, training hyper-box classifier can be mathematically formulated as Mixed Integer Linear Programming (MILP) predictions combine In this work, an optimisation-based approach proposed multi-class data using representation, thus facilitating extraction compact IF-THEN rules. The key novelty our lies minimisation number length generated rules enhanced Through real-world datasets, it demonstrated algorithm exhibits favorable performance when compared well-known alternatives prediction rule set simplicity.

Language: Английский

---

Deep transfer learning hybrid techniques for precision in breast cancer tumor histopathology classification

Health Information Science and Systems, Journal Year: 2025, Volume and Issue: 13(1)

Published: Feb. 11, 2025

Language: Английский

---