检索结果-内蒙古大学图书馆

3rd international conference on applied Intelligence and Informatics (AII)

作者： Alndiwee, Maher Preetha, V. K. Mathew, Jimy Palit, Shamik Joseph, Natalie datar, Akshay Lopez, Pablo Izquierdo Ullah, Waqar Univ Stirling RAK Campus Ras Al Khaymah U Arab Emirates RAK Municipal Rak U Arab Emirates RAK Res & Innovat Ctr Rak U Arab Emirates

ISBN: (纸本)9783031686382;9783031686399

This paper presents a comprehensive study on predicting the performance of outdoor comfort systems using machine learning (ML) models. The study is based on a dataset collected from various projects in Ras Al Khaimah (RAK), United Arab Emirates (UAE) and focuses on the prediction of temperature and humidity values. The evaluated models include Random Forest Learner, Neural Network, and Support Vector machine ( SVM), with performance assessed using key metrics. The findings of the study provide actionable insights from the data analysis process and the performance of machine learning models. Among the models evaluated, the Random Forest Learner demonstrated superior performance in capturing the variability in the data and generating accurate predictions. The Root Mean Square Error (RMSE) for temperature was 0.838 degrees C, and for humidity, it was 1.4%, which are considered relevant for estimating comfort levels by simulating the site's data before installing the cooling controls. This highlights its potential as a reliable tool for predicting outdoor comfort system performance. Furthermore, the study proposes integrating machine learning models into an enterprise-based decision support system for planning and managing outdoor comfort projects. By leveraging the predictive capabilities of these models, stakeholders can make informed decisions regarding system design and optimization, which helps them boost efficiency and reduce the cost of such costly projects.

关键词： Outdoor comfort system temperature prediction humidity prediction Random Forest Neural Network SVM data analysis machine learning IoT

来源：评论

学校读者我要写书评

暂无评论

Poisoning Attacks Against Non-IID Federated learning with Mixed-data Calibration 7

Poisoning Attacks Against Non-IID Federated Learning with Mi...

引用

7th international conference on machine learning and Natural Language Processing, MLNLP 2024

作者： Zhang, Xufei Uludag, Suleyman Laurel Springs School The Academy Ojai United States University of Michigan-Flint Department of Computer Science Flint United States

ISBN: (纸本)9798350354973

Federated learning (FL) is a privacy-preserving and collaborative machine learning approach that enables de-centralized data utilization across multiple clients. However, the performance of the global model can be compromised by the presence of non-independent and non-identically distributed (non-IID) data among clients. Various calibration methods, in-cluding Mixed data Calibration (MIDAC), have been proposed to address these challenges by improving the global model's accuracy. Despite their advantages, these methods introduce a new vulnerability: calibration data-poisoning attacks. To the best of our knowledge, our work is the first to study such attacks. This paper investigates the impact of calibration data-poisoning on FL systems using MIDAC. Through extensive experimentation with the CIFAR-10 dataset, we demonstrate that as the percentage of poisoned calibration data increases, the global model's accuracy significantly decreases. Furthermore, we show that increasing the size of the calibration dataset can mitigate some of the adverse effects of poisoned data. These findings highlight the urgent need for robust mechanisms to defend against calibration data-poisoning attacks, ensuring the security and reliability of FL systems using calibration methods. © 2024 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Integrating machine learning and Deep learning for Accurate Fraud Detection

Integrating Machine Learning and Deep Learning for Accurate ...

引用

2025 international conference on Artificial Intelligence and data Engineering, AIDE 2025

作者： Karthik, Morathoti Saigeeta, S.B. Reddy, P. Kushal Shiva Charith Reddy, Peddireddy Ram Dinesh Srinivas, M. Amrita Vishwa Vidhyapeetham Dept. of Computer Science Amaravati India Amrita Vishwa Vidyapeetham Dept. of Computer Science Andhra Pradesh Amaravati India

ISBN: (纸本)9798331527518

Credit card fraud detection is a major challenge for the financial industry, and its prevention requires machine learning models to be accurate and efficient in reducing financial losses. This paper proposes a comparative analysis of three most popular machine learning models DT, RF, and XGB that can be applied for fraud detection. To optimize the model's performance, we apply RFE to select only the most important features for the classification process. Each of the models has been evaluated for their performance through several metrics including accuracy, precision, recall, F1-score, and AUC-ROC, to validate its effectiveness on detecting fraudulent transactions. We have also analyzed each model's computational efficiency, not only in training but also for inference time because the real-time nature of this task is so crucial. While our results have shown that XGB outperformed DT and RF in both accuracy and AUC-ROC, it really does need fine hyperparameter tuning to optimize its computational cost. The results of the paper provide excellent insights for banks that want to deploy machine learning models for detecting fraud, by offering a compromise between model performance and computational efficiency. © 2025 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

data-Driven Earthquake Multi-impact Modeling:A Comparison of Models

引用

international Journal of Disaster Risk science 2024年第3期15卷 421-433页

作者： Hamish Patten Max Anderson Loake David Steinsaltz Information Management International Federation of Red Cross and Red Crescent Societies(IFRC)1211GenevaSwitzerland Department of Statistics University of OxfordOxfordOX13LBUK

In this study,a broad range of supervised machine learning and parametric statistical,geospatial,and non-geospatial models were applied to model both aggregated observed impact estimate data and satellite image-derived geolocated building damage data for earthquakes,via regression-and classification-based models,*** the aggregated observational data,models were ranked via predictive performance of mortality,population displacement,building damage,and building destruction for 375 observations across 161 earthquakes in 61 *** the satellite image-derived data,models were ranked via classification performance(damaged/unaff ected)of 369,813 geolocated buildings for 26 earthquakes in 15 *** k-fold,3-repeat cross validation was used to ensure out-of-sample predictive *** importance of several variables used as proxies for vulnerability to disasters indicates covariate *** 2023 Türkiye-Syria earthquake event was used to explore model limitations for extreme ***,applying the AdaBoost model on the 27,032 held-out buildings of the 2023 Türkiye-Syria earthquake event,predictions had an AUC of ***,without any geospatial,building-specific,or direct satellite image information,this model accurately classified building damage,with significantly improved performance over satellite image trained models found in the literature.

关键词： Disaster risk modeling Earthquake impact models machine learning Disaster statistics Satellite image-derived buildingdamage

来源：评论

学校读者我要写书评

暂无评论

Sensor fault detection, diagnosis, and data reconstruction strategy for chiller based on an improved tree-based pipeline optimization tool framework

引用

international JOURNAL OF REFRIGERATION 2025年 174卷 60-75页

作者： Wu, Pinguo Hu, Yunpeng Li, Guannan Liu, Qi Xiong, Chenglong Deng, Jiahui Chen, Shiao Wuhan Univ Sci & Technol Sch Urban Construct Wuhan 430065 Peoples R China Wuhan Business Univ Dept Bldg Environm & Energy Engn Wuhan 430065 Peoples R China Xian Univ Architecture & Technol State Key Lab Green Bldg Western China Xian 710055 Peoples R China Wuhan Univ Sci & Technol Hubei Prov Engn Res Ctr Urban Regenerat Wuhan 430081 Peoples R China China Railway SIYUAN Consulting & Engn Grp Co Ltd Wuhan 430063 Peoples R China

machine learning is widely applied to fault detection, diagnosis, and data reconstruction for chiller sensors but often requires domain expertise and manual intervention. Tree-based Pipeline Optimization Tool (TPOT), an automated machine learning framework, shows promise in fault detection, diagnosis, and reconstruction by automating model optimization and parameter tuning. Although the TPOT framework includes automated data preprocessing functions, it lacks the ability to automatically handle outliers. Outliers in sensor data can adversely affect the quality of the modeling process. By leveraging TPOT's capability for automated modeling, an ensemble fault diagnosis model can be developed. However, this model is prone to misdiagnosis when the sensor variables exhibit high correlations. Therefore, this study proposes an improved TPOT framework by incorporating a sliding window strategy to enhance TPOT's ability to handle outliers. The ensemble fault diagnosis model based on TPOT incorporates a Euclidean distance strategy, which identifies faulty sensors by quantifying the difference between the input data and the predicted results. Results show that the improved TPOT framework enhances fault detection, diagnosis, and data reconstruction. In the detection of sensor bias, drift, and precision degradation faults, the fault detection rates increased by a mean of 3.11 %, 4.64 %, and 8.62 %, respectively. The diagnostic strategy incorporating Euclidean distance reduced the number of misdiagnoses by one in the diagnosis of nine different sensor faults. In sensor data reconstruction, the RMSE was reduced by a mean of 68.26 %.

关键词： Chiller sensor Automated machine learning Tree-based Pipeline Optimization Tool Fault detection and diagnosis data reconstruction

来源：评论

学校读者我要写书评

暂无评论

A Big data, Bigger Impact: A Comprehensive Review of machine learning Advancements 4

A Big Data, Bigger Impact: A Comprehensive Review of Machine...

引用

4th international conference on Electrical Engineering and Computer science, ICECOS 2024

作者： Dlshad Hamad Ameen, Didar Wahhab Kareem, Shahab Baez Hasan, Sarhad Soran University Faculty of Science Computer Science Department Soran Kurdistan Iraq Engineering Erbil Polytechnic University Erbil Department of Information System Kurdistan Iraq

ISBN: (纸本)9798350368253

The exponential growth of data driven by the internet has necessitated effective extraction of insights, with big data and machine learning standing as pivotal tools. This paper aims to provide insights into the evolving landscape of machine learning and big data analytics through a comprehensive literature review. It seeks to underscore the key advancements and intersections between these domains and their transformative applications across industries. Specifically, the study will delve into the foundational theories that underpin machine learning algorithms, data collection and preprocessing techniques involved in big data analysis, and the prominent challenges associated with extracting insights from large and complex datasets. It will further explore major studies that exemplify the integration of machine learning methodologies and big data technologies for addressing real-world problems. The study will also examine visualization methods that facilitate comprehending insights gleaned from big data. Finally, it will discuss future research directions and highlight best practices for leveraging machine learning and big data analytics in decision-making processes. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Term Deposit Customer Prediction: A Blending Approach Using Bank Marketing data

Term Deposit Customer Prediction: A Blending Approach Using ...

引用

2025 IEEE international Students' conference on Electrical, Electronics and Computer science, SCEECS 2025

作者： Prasad, Rahul Choubey, Mohit Baghel, R.K. Naruka, Aaditya Upes Dehradun Cluster of Computer Science India Maulana Azad National Institute of Technology Dept. of Ece Bhopal India Poornima University Dept. of Se Jaipur India

ISBN: (纸本)9798331529833

The global COVID-19 pandemic led to a significant economic downturn, severely impacting financial systems and economies around the world. Widespread lockdowns and travel restrictions disrupted supply chains, forced business closures, and triggered job losses, highlighting the importance of sound financial planning. Term deposits have emerged as a preferred strategy for risk mitigation. In marketing, direct customer outreach has increasingly replaced traditional intermediary-based approaches. Using data from a Portuguese bank, this study explores the predictive accuracy of various machine learning models in identifying customers most likely to invest in term deposits. machine learning techniques applied include Support Vector machine (SVM), Gaussian Naïve Bayes, bagging, blending, Light Gradient Boosting machine (GBM), bagging, and Extreme Gradient Boosting (XGBoost). Results demonstrate that bagging algorithms provide superior accuracy in predicting customer behavior, offering insights to enhance targeted marketing efforts." © 2025 IEEE.

关键词： Sales

来源：评论

学校读者我要写书评

暂无评论

WAVE DYNAMICS RUN-UP MODELLING: machine learning APPROACH 43

WAVE DYNAMICS RUN-UP MODELLING: MACHINE LEARNING APPROACH

引用

ASME 43rd international conference on Ocean, Offshore and Arctic Engineering (OMAE)

作者： Vileti, Vinicius L. Ersdal, Svein Aker Solut Oslo Norway

ISBN: (纸本)9780791887837

This study proposes an alternative approach to predict wave elevation near multi-column semi-submersible structures by applying machine-learning methods from experimental data. The most common approach to this problem is to apply linear potential theory numerical programs to calculate the wave elevation close to such marine structures. However, in some cases, the assumptions for the potential theory are no longer valid, leading to a deviation from model test results. Another possible approach is to solve Navier Stokes EquationsNumerically (CFD), which poses a challenge regarding computational power for this kind of stochastic analysis. This paper details a procedure to apply machine learning to enhance the results by combining potential theory and experimental results for future predictions. Aker Solutions has performed experimental tests with a TLP (Tension Leg Platform) shaped hull under waves while measuring wave elevation on several points around it. These experimental data were treated and combined with potential theory results to compose a machine-learning prediction model. A frequency domain model was applied, where the experiment data is converted into the frequency domain and combined with the Potential Theory results to train the machine learning model. Results show that the machine learning model improves the results for wave elevations when closer to the hull, as those are the cases where the potential theory deviates more. This approach can also be applied to similar problems, such as wave loading and vessel motions.

关键词： Wave Dynamics Hydrodynamics Wave Loads machine learning AI Deep learning

来源：评论

学校读者我要写书评

暂无评论

A Novel Ensemble machine learning Framework for Improved Diabetes Prediction and Complication Prevention 3

A Novel Ensemble Machine Learning Framework for Improved Dia...

引用

3rd international conference on machine learning and data Engineering, ICMLDE 2024

作者： Jena, Siddhant Kumar Behera, Dayal Kumar Jena, Ajay Kumar Rout, Jitendra Kumar National Institute of Technology Odisha Rourkela India School of Computer Engineering KIIT Deemed to Be University Bhubaneswar India Department of Computer Science and Engineering National Institute of Technology Chhattisgarh Raipur India

This study introduces an innovative framework for predicting diabetes, employing advanced imputation strategies and ensemble machine learning techniques to boost prediction accuracy. Given the irreversible nature of diabetes, early diagnosis is paramount. We applied machine learning algorithms on the Pima Indian Diabetes (PID) dataset, comprising data from 768 patients with nine distinctive attributes, sourced from the UCI machine learning Repository. Our methodology integrates KNN Imputer for addressing missing data, SMOTE to tackle class imbalance, and RFE for optimal feature selection. We tested six machine learning algorithms: LR, DT, RF, SVM, Gradient Boosting (GB), and an ensemble Voting classifier. The ensemble model, which combines LR, DT, RF, SVM, and GB, outperformed individual models, achieving a top accuracy of 84.2%, with a root mean squared error (RMSE) of 0.334 and an R2 score of 0.528. This study highlights the enhanced predictive power achievable through the synergy of multiple machine learning models. © 2025 The Authors. Published by Elsevier B.V.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

Criminal Activity Forecasting using machine learning 10

Criminal Activity Forecasting using Machine Learning

引用

10th international conference on Advanced Computing and Communication Systems, ICACCS 2024

作者： Augustin Sam, N. Bamini, A. M. Anusha Karunya Institute of Technology and Sciences Division of Computer Science and Engineering Tamil Nadu Coimbatore India

ISBN: (纸本)9798350384369

Crime analysis is an important topic that uses cutting-edge machine learning. This study focuses on using machine learning algorithms to examine crime trends and give law enforcement organizations useful information. The ML models may find patterns and trends by using past crime data, which makes it possible to identify possible crime hotspots and allocate the necessary resources. To find hidden patterns in crime data, Random forest algorithm was applied. The algorithm has the ability to automatically group criminal episodes into clusters, categorize them according to various attributes, and spot anomalies-that is, deviations from the norm. Furthermore, by using previous data, machine learning algorithms can be trained to support proactive crime prevention tactics. The accuracy and efficacy of crime analysis models are further improved by the integration of many data sources, such as information. The proposed algorithm makes it easier to monitor crimes in real time and permits prompt decision-making in emergency scenarios. Law enforcement organizations can use their resources more wisely, plan out patrol routes, and create focused crime prevention programs. By giving law enforcement agencies useful insights, machine learning turns crime analysis into a potent weapon. © 2024 IEEE.

关键词： Adversarial machine learning

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：