An advanced performance-based method for soft and abrupt fault diagnosis of industrial gas turbines

Energy, Journal Year: 2025, Volume and Issue: unknown, P. 135358 - 135358

Published: March 1, 2025

Language: Английский

---

Comparison and Analysis of Multiple Machine Learning Algorithms for Predicting Student Adaptation Levels in Online Education

Lecture Notes in Education Psychology and Public Media, Journal Year: 2024, Volume and Issue: 40(1), P. 30 - 37

Published: March 4, 2024

With the rapid development and popularization of Internet technology, online education has become a new way education. Compared with traditional classroom teaching, more flexible learning mode, convenient environment wider range resources. However, at same time, also faces some challenges, one most important challenges is adaptability students to In this paper, we use machine techniques predict students' in classrooms. After using logistic regression model, k-neighborhood algorithm random forest XGBoost model Cat Boost make predictions, it found that best predicting classroom, prediction accuracy 89.6%. The CatBoost were better prediction, accuracies 89.1% 88.6%, respectively. contrast, KNN models have poorer 68.8% 77.1%, research article implications for industry. By an can help educational institutions understand improve teaching effectiveness. Meanwhile, students, knowing their adaptive ability helps them plan study programs efficiency. This uses classrooms, results show performs terms predictive provides useful reference industry ideas future research.

Language: Английский

---

Comparison and Analysis of the Accuracy of Various Machine Learning Algorithms in Bitcoin Price Prediction

Advances in Economics Management and Political Sciences, Journal Year: 2024, Volume and Issue: 70(1), P. 302 - 308

Published: Jan. 5, 2024

Based on the dataset of Bitcoin Price dataset, this paper studied price prediction by using support vector machine model, random forest neural network XGBoost model and LightGBM model. The models were evaluated MSE, RMSE, MAE, MAPE R. First, we divided into a training set test according to ratio 7:3, with 70 as 30 set. We take stock change (return) target variable, other variables input variables, use train After comparison, found that XGBoost's R are all optimal, its effect is also best. performance four ranges from good different, including LightGBM, Forest, network. Among them, MSE dozens times models, so it performs worst. well in dealing high-dimensional sparse data nonlinear relationships, while Forest suitable for large-scale data. Support machines networks require more tuning optimization advantage their advantages. In summary, research results can provide value future, certain reference selecting learning

Language: Английский

---

A multi-objective optimization model for the location of cold chain logistics distribution center based on trust domain optimization algorithm

Applied and Computational Engineering, Journal Year: 2024, Volume and Issue: 50(1), P. 183 - 188

Published: March 22, 2024

Cold chain transportation refers to a logistics method that transports fresh, perishable, and perishable items from the place of production, processing or warehouse consumption under certain temperature conditions. As consumers have higher requirements for food safety quality, cold industry has also developed rapidly. However, high cost logistics, difficulty technology, ensuring service quality other problems arisen, how optimize become hot issue in industry. optimization premise goods, through reasonable paths, methods, control measures, minimize transportation, improve efficiency quality. In order achieve it is necessary establish reliable distribution center ensure goods during transportation. solve problem location selection center, this paper first defines reliability calculation center. Then, model trust domain algorithm were established. The aims total cost, while considering multiple factors such as distance, control, facility construction, etc., operational Finally, introduces an example prove feasibility universality Through examples, can be seen effectively reduce summary, proposes centers challenges faced by This provide guidance companies gain greater advantage highly competitive market.

Language: Английский

---

Role of education and awareness programs in fostering energy conservation behavior in cities: Empowering urban sustainability using deep learning approach

Sustainable Cities and Society, Journal Year: 2024, Volume and Issue: 112, P. 105505 - 105505

Published: May 6, 2024

Language: Английский

---

Comparison and Analysis of the Effect of XGBoost Classification, BP Neural Network Classification and CatBoost Classification on Malware Attack Prediction

Published: Nov. 24, 2023

Language: Английский

---

Analysis of annual average daily concentration of PM2.5 particles in Urumqi and prediction of average daily concentration in the next 5 years based on time series model

Published: Feb. 21, 2024

Language: Английский

---

Overview of Flexible Load Control

Power systems, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 8

Published: Jan. 1, 2024

Language: Английский

---

Real-Time Short-Circuit Current Calculation in Electrical Distribution Systems Considering the Uncertainty of Renewable Resources and Electricity Loads

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(23), P. 11001 - 11001

Published: Nov. 26, 2024

Existing short-circuit calculation methods for distribution networks with renewable energy sources ignore the fluctuation of and cannot reflect impact load changes on current in real time at all times day extreme scenarios. A real-time method is proposed to take into account stochastic nature distributed generators (DGs) electricity loads. Firstly, continuous power flow calculated based output And then, equivalent DG models low-voltage ride through (LVRT) strategies are substituted iterative obtain currents main branches time. The effects different curves network quantitatively analyzed during output, which can provide an important basis setting relay protection study new principles protection.

Language: Английский

---