Predicting employee attrition and explaining its determinants DOI Creative Commons
Shahin Manafi Varkiani, Francesco Pattarin, Tommaso Fabbri

и другие.

Expert Systems with Applications, Год журнала: 2025, Номер unknown, С. 126575 - 126575

Опубликована: Янв. 1, 2025

Язык: Английский

Quantifying compressive strength in limestone powder incorporated concrete with incorporating various machine learning algorithms with SHAP analysis DOI

Mihir Mishra

Asian Journal of Civil Engineering, Год журнала: 2024, Номер unknown

Опубликована: Дек. 14, 2024

Язык: Английский

Процитировано

11

Enhancing wind power prediction with self-attentive variational autoencoders: A comparative study DOI Creative Commons
Fouzi Harrou, Abdelkader Dairi, Abdelhakim Dorbane

и другие.

Results in Engineering, Год журнала: 2024, Номер 23, С. 102504 - 102504

Опубликована: Июль 14, 2024

Accurate wind power prediction is critical for efficient grid management and the integration of renewable energy sources into grid. This study presents an effective deep-learning approach that improves short-term forecasting accuracy. The method incorporates a Variational Autoencoder (VAE) with self-attention mechanism applied in both encoder decoder. empowers model to leverage VAE's strengths time-series modeling nonlinear approximation while focusing on most relevant features within data. effectiveness this evaluated through comprehensive comparison eight established deep learning methods, including Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM) networks, Bidirectional LSTMs (BiLSTMs), Convolutional (ConvLSTMs), Gated Units (GRUs), Stacked Autoencoders (SAEs), Restricted Boltzmann Machines (RBMs), vanilla VAEs. Real-world data from five turbines France Turkey used evaluation. Five statistical metrics are employed quantitatively assess performance each method. results indicate SA-VAE consistently outperformed other models, achieving highest average R2 value 0.992, demonstrating its superior predictive capability compared existing techniques.

Язык: Английский

Процитировано

10

Advancing forest fire prediction: A multi-layer stacking ensemble model approach DOI

Fahad Shahzad,

Kaleem Mehmood, Shoaib Ahmad Anees

и другие.

Earth Science Informatics, Год журнала: 2025, Номер 18(3)

Опубликована: Фев. 19, 2025

Язык: Английский

Процитировано

2

Prediction of microalgae harvesting efficiency and identification of important parameters for ballasted flotation using an optimized machine learning model DOI
Kaiwei Xu, Zihan Zhu, Haining Yu

и другие.

Algal Research, Год журнала: 2025, Номер unknown, С. 103985 - 103985

Опубликована: Фев. 1, 2025

Язык: Английский

Процитировано

2

Algorithms and tools for data-driven omics integration to achieve multilayer biological insights: a narrative review DOI Creative Commons
Aurelia Morabito, Giulia De Simone, Roberta Pastorelli

и другие.

Journal of Translational Medicine, Год журнала: 2025, Номер 23(1)

Опубликована: Апрель 10, 2025

Systems biology is a holistic approach to biological sciences that combines experimental and computational strategies, aimed at integrating information from different scales of processes unravel pathophysiological mechanisms behaviours. In this scenario, high-throughput technologies have been playing major role in providing huge amounts omics data, whose integration would offer unprecedented possibilities gaining insights on diseases identifying potential biomarkers. the present review, we focus strategies applied literature integrate genomics, transcriptomics, proteomics, metabolomics year range 2018-2024. Integration approaches were divided into three main categories: statistical-based approaches, multivariate methods, machine learning/artificial intelligence techniques. Among them, statistical (mainly based correlation) ones with slightly higher prevalence, followed by learning Integrating multiple layers has shown great uncovering molecular mechanisms, putative biomarkers, aid classification, most time resulting better performances when compared single analyses. However, significant challenges remain. The nature platforms introduces issues such as variable data quality, missing values, collinearity, dimensionality. These further increase combining datasets, complexity heterogeneity integration. We report found cope these challenges, but some open still remain should be addressed disclose full

Язык: Английский

Процитировано

2

Revealing the effects of environmental and spatio-temporal variables on changes in Japanese sardine (Sardinops melanostictus) high abundance fishing grounds based on interpretable machine learning approach DOI Creative Commons
Yongchuang Shi,

Lei Yan,

Shengmao Zhang

и другие.

Frontiers in Marine Science, Год журнала: 2025, Номер 11

Опубликована: Янв. 13, 2025

The construction of accurate and interpretable predictive model for high abundance fishing ground is conducive to better sustainable fisheries production carbon reduction. This article used refined statistical maps visualize the spatial temporal patterns catch changes based on 2014-2021 fishery statistics Japanese sardine Sardinops melanostictus in Northwest Pacific Ocean. Three models (XGBoost, LightGBM, CatBoost) two variable importance visualization methods (model built-in (split) SHAP methods) were comparative analysis determine optimal modeling strategies. Results: 1) From 2014 2021, annual showed an overall increasing trend peaked at 220,009.063 tons 2021; total monthly increased then decreased, with a peak 76, 033.4944 (July), was mainly concentrated regions 39.5°-43°N 146.75°-155.75°E; 2) Catboost predicted than LightGBM XGBoost models, highest values accuracy F1-score, 73.8% 75.31%, respectively; 3) ranking model’s method differed significantly from that method, variables increased. Compared informs magnitude direction influence each global local levels. results research help us select construct prediction grounds Ocean, which will provide scientific basis achieve environmental economically development.

Язык: Английский

Процитировано

1

Overview of Data-Driven Models for Wind Turbine Wake Flows DOI Creative Commons
Maokun Ye, Min Li,

Mingqiu Liu

и другие.

Journal of Marine Science and Application, Год журнала: 2025, Номер unknown

Опубликована: Фев. 8, 2025

Язык: Английский

Процитировано

1

An Explainable XGBoost Model for International Roughness Index Prediction and Key Factor Identification DOI Creative Commons

Bin Lv,

Haixia Gong,

Dong Bin

и другие.

Applied Sciences, Год журнала: 2025, Номер 15(4), С. 1893 - 1893

Опубликована: Фев. 12, 2025

This study proposes an explainable extreme gradient boosting (XGBoost) model for predicting the international roughness index (IRI) and identifying key influencing factors. A comprehensive dataset integrating multiple data sources, such as structure, climate traffic load, is constructed. voting-based feature selection strategy adopted to identify factors, which are used inputs prediction model. Multiple machine learning (ML) models trained predict IRI with constructed dataset, XGBoost performs best coefficient of determination (R2) reaching 0.778. Finally, interpretable techniques including importance, Shapley additive explanations (SHAP) partial dependency plots (PDPs) employed reveal mechanism factors on IRI. The results demonstrate that conditions load play a critical role in deterioration provides relatively universal perspective factor identification, outputs proposed method contribute making scientific maintenance strategies roads some extent.

Язык: Английский

Процитировано

1

Sustainable AI-driven wind energy forecasting: advancing zero-carbon cities and environmental computation DOI Creative Commons
Haytham H. Elmousalami,

Aljawharah A. Alnaser,

Felix Kin Peng Hui

и другие.

Artificial Intelligence Review, Год журнала: 2025, Номер 58(6)

Опубликована: Март 29, 2025

Язык: Английский

Процитировано

1

Unraveling volatile metabolites in pigmented onion (Allium cepa L.) bulbs through HS-SPME/GC–MS-based metabolomics and machine learning DOI Creative Commons

Kaiqi Cheng,

Jin-Chang Xiao,

Jingyuan He

и другие.

Frontiers in Nutrition, Год журнала: 2025, Номер 12

Опубликована: Апрель 22, 2025

Colored onions are favored by consumers due to their distinctive aroma, rich phytochemical content, and diverse biological activities. However, comprehensive analyses of profiles volatile metabolites remain limited. In this study, total phenols, flavonoids, anthocyanins, carotenoids, antioxidant activities three colored onion bulbs were evaluated. Volatile identified using headspace solid-phase microextraction combined with gas chromatography-mass spectrometry (HS-SPME/GC-MS). Multivariate statistical analyses, feature selection techniques (SelectKBest, LASSO), machine learning models applied further analyze classify the metabolite profiles. Significant differences in composition observed among types. A 243 detected, sulfur compounds accounting for 51-64%, followed organic acids derivatives (4-19%). analysis revealed distinct profiles, 19 key as biomarkers. Additionally, 33 38 selected SelectKBest LASSO, respectively. The features LASSO enabled clear differentiation types via PCA, UMAP, k-means clustering. Among four tested, random forest model achieved highest classification accuracy (1.00). SHAP confirmed 20 potential markers. findings suggest that combination HS-SPME/GC-MS learning, particularly algorithm, is a powerful approach characterizing classifying onions. This method holds quality assessment breeding applications.

Язык: Английский

Процитировано

1