Enhancing Thermal Comfort and Energy Efficiency in Building Environments Using Advanced Dynamic Bayesian Modeling

Journal of Building Engineering, Год журнала: 2025, Номер unknown, С. 112951 - 112951

Опубликована: Май 1, 2025

Язык: Английский

Research on the prediction model of mastitis in dairy cows based on time series characteristics

Frontiers in Veterinary Science, Год журнала: 2025, Номер 12

Опубликована: Апрель 24, 2025

Mastitis in dairy cows is a significant challenge faced by the global industry, significantly affecting quality and output of milk from enterprises causing them to suffer severe economic losses. With increasing public concern over food safety rational use antibiotics, how identify at risk disease early has become key issue that needs be urgently addressed. Especially subclinical mastitis, due lack obvious external symptoms, makes detection more difficult, so warning it particularly important. In this study, time series prediction method, combined with machine learning techniques, was used predict mastitis cows. The study data were obtained production records 4000 large farm Hexi region Gansu. By constructing time-series features, indicators such as yield, fat rate protein each cow two consecutive months, April May, utilized its health status June. To fully exploit value we designed multidimensional feature set included raw indicator values, monthly change rates, statistical features. After preprocessing sample balancing, 2821 selected for model training. Finally, applicability assessed comparing analyzing performance six models, namely eXtreme Gradient Boosting(XGBoost), Boosting Decision Tree (GBDT), Support Vector Machine (SVM), K Nearest Neighbors (KNN), Logistic Regression, Long Short-Term Memory Network (LSTM). XGBoost demonstrated optimal performance, achieving an area under ROC curve (AUC) 0.75 accuracy 71.36%. Feature importance analysis revealed three temporal influencing outcomes: May yield (22.29%), standard deviation percentage (20.27%), (19.87%). SHapley Additive exPlanations (SHAP) further validated predictive these providing managers clearly defined monitoring priorities. demonstrates strong potential accurate tool This presents effective early-warning approach through modeling offers practical prevention management.

Язык: Английский

Research on recognition of bedding system coverage rate using infrared thermal imaging

Journal of Thermal Biology, Год журнала: 2025, Номер 130, С. 104146 - 104146

Опубликована: Май 1, 2025

Язык: Английский