检索结果-内蒙古大学图书馆

Credit Card Fraud Detection Based on Hyperparameters Optimization Using the Differential Evolution

INTERNATIONAL JOURNAL OF INFORMATION SECURITY AND PRIVACY 2022年第1期16卷

作者： Tayebi, Mohammed El Kafhali, Said Hassan First Univ Settat Fac Sci & Tech IR2M Lab Settat Morocco

Due to the emigration of world business to the internet, credit cards have become a tool for payments for both online and outline purchases. However, fraudsters try to attack those systems using various techniques, and credit card fraud has become dangerous. To secure credit cards, different methods are proposed in the academic paper based on artificial intelligence. The proposed solution in this paper aims at combining the robustness of three methods: the differential evolution algorithm (DE) for selecting the best hyperparameters, a resampling technique for handling imbalanced data issues, and the xgboost technique for classification. Finally, the fraudulent transactions are classified using the optimized xgboost algorithm. The proposed solution is evaluated using two real-world datasets: the European dataset and the UCI dataset. The evaluation n terms of accuracy, sensitivity, specificity, precision, and F-measure shows the ability and the superiority of the proposed approach in comparison with the state-of-the-art machine learning models.

关键词： Differential Evolution algorithm Fraud Transactions Hyperparameter Optimization Resampling Techniques xgboost algorithm

来源：评论

学校读者我要写书评

暂无评论

A hybrid machine learning approach for early mortality prediction of ICU patients

引用

PROGRESS IN ARTIFICIAL INTELLIGENCE 2022年第4期11卷 333-347页

作者： Mansouri, Ardeshir Noei, Mohammadreza Abadeh, Mohammad Saniee Tarbiat Modares Univ Dept Elect & Comp Engn Tehran Iran George Mason Univ Dept Comp Sci Fairfax VA 22030 USA

Hospitals face many pressures, including limited budgets and resources. The intensive care unit (ICU), has attracted much attention from the medical community. Patients in the ICU are monitored continuously and their physiological data is consistently measured. This provides a valuable opportunity to analyze significant clinical data. Predicting the mortality of patients in the ICU is crucial. By using this prediction, intensivists identify patients who will not benefit from receiving treatment in the ICU and focus more on the care of patients who will benefit from this treatment. To date, various scoring systems and machine learning models have been developed to predict mortality in intensive care units. In this paper, our goal is to provide a model that can predict mortality for up to 24 h after ICU admission. This research has been conducted using Medical Information Mart for Intensive Care III (MIMIC-III) database. Relevant data have been extracted, preprocessed, and prepared for data mining analysis. The proposed framework has two algorithmic stages. In the first step, time-series data within 24 h of admission is fed to a convolutional neural network with optimized hyperparameters. In the second step, by using the output of the previous stage and a filtering strategy for temporal features, a new data set is created and is fed to an xgboost algorithm, for final classification. The proposed framework outperforms severity of illness scores and machine learning models within 24 h of admission to the ICU and attains a ROC AUC of 0.863 (+/- 0.004).

关键词： Mortality prediction ICU MIMIC III Convolutional neural network xgboost algorithm Hybrid approach

来源：评论

学校读者我要写书评

暂无评论

Experimental Study and Machine Learning Aided Modelling of the Mechanical Behaviour of Rammed Earth

引用

GEOTECHNICAL AND GEOLOGICAL ENGINEERING 2022年第10期40卷 5007-5027页

作者： Kardani, Navid Zhou, Annan Lin, Xiaoshan Nazem, Majidreza Royal Melbourne Inst Technol RMIT Sch Engn Civil & Infrastruct Engn Discipline Melbourne Vic 3001 Australia

Rammed earth is a sustainable building technique for constructing foundations, floors, and walls using natural raw materials such as earth, chalk, lime, with stabilizers like cements. As the proportion of various materials changes, the mechanical properties of rammed earth materials are also varying correspondingly. A series of experimental studies are first conducted to evaluate the effects of different proportions of raw materials including clay, sand, cement, and water under various loading rates on the strength/deformation properties (peak strength, q(f);residual strength, q(res);initial modulus, E-max;secant modulus at 50% peak strength, E-50) and stress-strain relationships (sigma(1) similar to epsilon(1)) of rammed earth. A soft computing method (extreme gradient boosting machine, xgboost) is then developed to model peak strength, residual strength, initial modulus, secant modulus and entire stress-strain relationships obtained from the experimental studies. Three performance metrics including the root mean squared error, variance accounted for and R-squared value (R-2) are used to measure the performance of the applied model. Comparisons between simulations and experiments show that the developed xgboost algorithm is a promising alternative in modelling key mechanical properties and entire stress-strain relationships for rammed earth. For stress-strain relationships calculated R-squared value for the training set is 0.978 and that for the testing dataset is 0.908. The key factor that most significantly affects the peak strength, residual strength, initial modulus, secant modulus and entire stress-strain relationships for rammed earth can be identified by using the developed soft computing method.

关键词： Rammed earth Key mechanical properties Peak strength Initial modulus xgboost algorithm Experimental analysis

来源：评论

学校读者我要写书评

暂无评论

Dynamic Forecasting of Solar Energy Microgrid Systems Using Feature Engineering

引用

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS 2022年第6期58卷 7857-7869页

作者： Mohamed, Muamar Mahmood, Farhad E. E. Abd, Mehmmood A. A. Chandra, Ambrish Singh, Bhim Univ Quebec Ecole Technol Super Dept Elect Engn Montreal PQ H3C 1K3 Canada Univ Mosul Dept Elect Engn Mosul 00964 Iraq IBM Corp Austin TX 78758 USA Indian Inst Technol Delhi New Delhi 110016 India

The intermittent and stochastic nature of solar energy generation systems, climate change, and the inefficiency of modern power systems due to zero inertia have created many challenges for on-grid operators. Solar forecasting systems based on machine learning algorithms are an emerging and effective solution that uses Big Data related to weather phenomena. However, the predictive ability of these algorithms is hampered by the sporadic nature of solar energy generation. In this article, a robust hybrid machine learning system that utilizes multiple linear regression (MLR) and a Pearson correlation coefficient (PCC) was tested on solar power plant sites of varying capacities in Germany (100-8500 kW). The volume of Big Data features can be reduced by focusing on the features that significantly improve the reliability of the mid-term forecasting system. In this way, drastic fluctuations in the prediction of photovoltaic (PV) power generation can be avoided. The results of our approach are evaluated regarding real-world data using the extreme gradient boosting (xgboost) with feature engineering, and principal component analysis (PCA), in order to forecast PV energy, rank, and track the importance of feature engineering for different PV capacities. Furthermore, we found that the need for selectivity and reduction of performance error was supported by ridge regression. In addition, the proposed novel xgboost forecast system decreased the root-mean-square error (RMSE) and mean absolute error (MAE) by 30% and 18%, respectively, compared to the Autoencoder and long short-term memory (LSTM) for same datasets. Furthermore, the CoD determination coefficient (R-2) increased by 85% compared to the statistical model's autoregressive integrated moving average (ARIMA).

关键词： Artificial intelligence data mining algorithms forecasting techniques machine learning power system statistical analysis solar microgrid xgboost algorithm

来源：评论

学校读者我要写书评

暂无评论

Photovoltaic Power Prediction Model Using Pre-train and Fine-tune Paradigm Based on LightGBM and xgboost 18

Photovoltaic Power Prediction Model Using Pre-train and Fine...

引用

18th International Conference on Future Networks and Communications (FNC) / 20th International Conference on Mobile Systems and Pervasive Computing (MobiSPC) / 13th International Conference on Sustainable Energy Information Technology (SEIT)

作者： Ye, Jiandong Zhao, Binqi Deng, Hua Guangdong Power Grid Co Ltd Yangjiang Power Supply Bur Yangjiang Peoples R China Inst Trop & Marine Meteorol Guangzhou Peoples R China

Accurate photovoltaic power prediction is crucial for optimizing the performance of photovoltaic power plants and ensuring the stability of the power grid. However, existing forecasting methods are challenged when predicting the power output of multiple plants with different geographical locations and environmental conditions simultaneously, due to the complexity of the underlying factors that affect power generation. To address this issue, this paper proposes a novel LightGBM-xgboost photovoltaic power prediction model based on a pre-training and fine-tuning paradigm. The model leverages a large dataset to pre-train and learn rich feature representations within meteorological elements, which are then fine-tuned on specific downstream tasks to improve prediction accuracy and generalization ability. Experimental verification is conducted using data from three photovoltaic generators operating in Guangdong, China. Results demonstrate that the proposed model outperforms other methods when predicting the power output of multiple plants with different conditions simultaneously, highlighting the effectiveness of the proposed approach. Overall, this work contributes to advancing the state-of-the-art in photovoltaic power prediction and has practical implications for improving the control and scheduling of photovoltaic power plants, as well as ensuring the stability and security of the power grid.

关键词： Photovoltaic power prediction Pre-train and Fine-tune model LightGBM algorithm xgboost algorithm

来源：评论

学校读者我要写书评

暂无评论

Multi-factor offshore short-term wind power prediction based on xgboost 35

Multi-factor offshore short-term wind power prediction based...

引用

35th Chinese Control and Decision Conference (CCDC)

作者： Zhang, Yuhan Wang, Bin Xu, Wanwan Wuhan Univ Sci & Technol Wuhan Peoples R China Wuhan Univ Sci & Technol Sch Informat Sci & Engn Wuhan Peoples R China

ISBN: (纸本)9798350334722

Fast and accurate wind power prediction is important for grid-connected operation and grid dispatch of offshore wind power. A gated recurrent neural network (GRU) multi-factor prediction model based on extreme gradient boosting (xgboost) algorithm and particle swarm optimization is proposed to address the problems of random fluctuation of offshore wind power, susceptibility to climate factors, and poor prediction accuracy due to large data volume. The relevance and influence of historical offshore wind power, wind speed, wind direction and temperature are analyzed, ranked in order of importance and downscaled, and then predicted using an optimized gated recurrent neural network. The results are validated with real data from an offshore wind farm and show a decrease in MAE, RMSE and SMAPE. The results show that the MAE, RMSE and SMAPE are all reduced, and that the MAE is still highly accurate when the offshore wind farm is in a different season or in a different location, thus demonstrating the adaptability of the model.

关键词： offshore wind power power prediction gated cycle unit xgboost algorithm neural networks

来源：评论

学校读者我要写书评

暂无评论

Comparative analysis of data using machine learning algorithms: A hydroponics system use case

引用

SMART AGRICULTURAL TECHNOLOGY 2023年 4卷

作者： Idoje, Godwin Mouroutoglou, Christos Dagiuklas, Tasos Kotsiras, Anastasios Muddesar, Iqbal Alefragkis, Panagiotis London South Bank Univ Comp Sci & Informat Div London England Univ Peloponnese Dept Agr Kalamata Greece Univ Peloponnese Dept Elect & Comp Engn Kalamata Greece London South Bank Univ Comp Sci 103 Borough Rd London SE1 0AA England

This paper makes a comparison of machine learning algorithms for the analysis of four hydroponic datasets. Data have been gathered daily from hydroponic systems to predict the output of the hydroponic systems. This research compares the performance of the federated split Learning, Deep neural network, extreme Gradient Boosting (xgboost), and Linear regression algorithms on four different hydroponic systems. These algorithms have been used to analyze the datasets of Nutrient Film Technic (NFT), Floating (FL), Aggregate (AG) and Aeroponic (AER) hydroponic systems. The results have indicated the performance of each model for each hydroponic system and how each algorithm have used the various multiple input features to make predictions of the onion bulb diameter and the errors encountered by each model. From the results obtained, it has been observed that the R square score is varied for each hydroponic system. This variation in the result has been also reflected in the Mean absolute errors obtained. This research determines which of the algorithms predict the optimal Onion bulb diameter (mm) using days after transplant (days), Temperature (degrees C), water consumption (Litres), Number of Leaves (NL), Nitrogen (mg/g), Phosphorus (mg/g), Potassium (mg/g), Calcium (mg/g), Magnesium (mg/g), Sulphur (mg/g), Sodium (mg/g) as independent variables. The results will be a guide in the choice of hydroponic system to adopt for food production based on the climatic parameters of the location, which is one of the numerous contributions of this research.

关键词： Federated split learning Nutrient Film Technic Aggregate Aeroponics Floating hydroponics Mean absolute error Regression xgboost algorithm

来源：评论

学校读者我要写书评

暂无评论

Research on loyalty prediction of e-commerce customer based on data mining

引用

APPLIED MATHEMATICS AND NONLINEAR SCIENCES 2022年第2期8卷 721-732页

作者： Qin, Xujie Anhui Business & Technol Coll Hefei 231131 Anhui Peoples R China

Analysing the data generated by the daily operations of enterprises through data mining technology can effectively predict customer loyalty and help enterprise leaders make correct decisions. Therefore, this paper classifies and analyses churn and loyalty of e-commerce customers, and by combining the application foundation of data mining technology in e-commerce customer loyalty prediction, a prediction model of e-commerce customer loyalty based on data mining is constructed. In this model, the local abnormal factor algorithm is used to eliminate the data for cleaning, the xgboost algorithm is improved by adding penalty coefficient, and the prediction effect of the model is evaluated and compared according to the values of Accuracy, Precision, Recall and F. The results show that the model has high accuracy in predicting customer loyalty, which can accurately extract attributes of users and characteristic information of commodities.

关键词： Data mining e-commerce customer loyalty xgboost algorithm

来源：评论

学校读者我要写书评

暂无评论

Load Demand Forecasting in Smart Microgrid Using NWP via Feature Engineering

Load Demand Forecasting in Smart Microgrid Using NWP via Fea...

引用

2023 IEEE International Conference on Energy Technologies for Future Grids, ETFG 2023

作者： Mohamed, Muamar Mahmood, Farhad E. Abd, Mehmmood A. Rezkallah, Miloud Hamadi, Abdelhamid Chandra, Ambrish Ecole de Technologie Superieure University of Quebec Department of Electrical Engineering MontrealQC Canada University of Mosul Department of Electrical Engineering Mosul Iraq Electrical and Computer Engineering Ibm United States

ISBN: (纸本)9781665471640

This study developed a machine learning model to forecast electrical demand for smart micro-grids with the existing of numerical weather forecasting (NWP). The model uses three techniques, linear regression (LR), Long Short-Term Memory (LSTM) and eXtreme Gradient Boosting (xgboost), to harness insights from historical data and predict future load patterns. Notably, the xgboost algorithm is highlighted as the most effective due to underscoring the importance of feature selection and model optimization. The utilization of these techniques results in accurate electrical demand predictions, which are crucial for enhancing micro-grid efficiency, practically in response to weather-induced load fluctuations. This paper suggests employing advanced feature selection methods based on high feature importance, ultimately refining the model's predictive capabilities for smart micro-grid applications. © 2023 IEEE.

关键词： forecasting techniques Load demand machine learning micro-grid NWP xgboost algorithm

来源：评论

学校读者我要写书评

暂无评论

The Implementation of Integration of ML algorithm for the Design of Stability System for Grid System

The Implementation of Integration of ML Algorithm for the De...

引用

2023 International Conference on Power Energy, Environment and Intelligent Control, PEEIC 2023

作者： Chandani, Priyanka Vijay, Sandip Soumya, K. Noida Institute of Engineering and Technology Department of Data Science Uttar Pradesh Greater Noida India Tula's Institute Department of Computer Science and Engineering Dehradun India Department of Computer Science Bangalore India

ISBN: (纸本)9798350357769

A modern decentralized electric grid is a groundbreaking system that integrates demand response effortlessly and doesn't need major infrastructure changes. Within the decentralized domain, users independently control their power consumption according to the frequency of the grid. This is made possible by the use of reasonably priced devices like smart meters, which allow grid frequency to be measured from almost anywhere. Different data-level resampling strategies have been used to address the problem of data imbalance, while data normalization approaches have been used to reduce biased behavior among characteristics. The findings clearly show that in terms of the performance of classifiers, a balanced dataset performs better than an unbalanced one. Specifically, for unbalanced datasets, oversampling approaches are more effective than under sampling ones. With a precision level of 94.7 percent, the xgboost algorithm is the best performer within the range of deep learning algorithms that are taken into consideration. Remarkably, xgboost's accuracy forecast rises to 96.8% when paired with random oversampling. This improved model manages the volatility of renewable energy supplies and maximizes their use by accurately forecasting frequency variations in decentralized power networks. This model's predictive capacities have great potential to support the stability of distributed electricity grids, which will improve the distribution and administration of energy on a larger scale. © 2023 IEEE.

关键词： Data Normalization Data Resampling Decentralized Power Grid Demand Response and Classifiers Performance Energy Policy Grid Stability Machine Learning Renewable Energy Integration Renewable Sources Smart Meters Stability Prediction xgboost algorithm

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：