检索结果-内蒙古大学图书馆

Computers, Information processing and Advanced Education (CIPAE), 2020 international conference on

作者： Liao Sha College of Art and Design Xiangnan University Chenzhou China

ISBN: (数字)9798331527662

ISBN: (纸本)9798331527679

This paper systematically expounds the implementation path and case application of algorithm-driven brand design. Firstly, it introduces the process of establishing a multi-channel data collection mechanism to obtain massive user behavior data and preprocess it to form a high-quality dataset. Secondly, it illustrates the key technologies for designing generation and optimization based on preprocessed data, such as dual-path convolutional neural networks, generative adversarial networks, etc. Furthermore, it elaborates on the method of integrating algorithms with existing design processes to achieve human-machine collaboration and enhance decision-making efficiency. Finally, it discusses the importance of continuously optimizing algorithm models through “online learning” mechanisms to ensure their up-to-dateness. The paper inserts multiple real cases and supplements them with chart data, comprehensively demonstrating the innovative applications of algorithm-driven brand design in multiple fields. Overall, algorithms are reshaping a new paradigm of brand design, providing unprecedented personalized experiences for consumers.

关键词： Reviews Human-machine systems Force Education Decision making Information processing Generative adversarial networks data models Convolutional neural networks Optimization

来源：评论

学校读者我要写书评

暂无评论

Research on data-Driven Financial Risk Prediction Model Based on Cloud Computing

Research on Data-Driven Financial Risk Prediction Model Base...

引用

Computers, Information processing and Advanced Education (CIPAE), 2020 international conference on

作者： Li Qinyi Communication University of China Nanjing China

ISBN: (数字)9798331527662

ISBN: (纸本)9798331527679

This article aims to explore a data-driven financial risk prediction model based on cloud computing technology. This model achieves accurate identification and prediction of corporate financial risks by integrating advanced algorithms and knowledge graph technology. This article proposes a new risk prediction algorithm model that combines multiple methods such as machine learning, deep learning, and statistical analysis to learn patterns from historical data and predict future financial risks. In addition, the application of knowledge graph enables the model to better understand the internal connections of financial data and improve the accuracy of predictions. Finally, through system simulation experiments, this paper verifies the effectiveness and stability of the proposed model in practical applications. Experimental results show that this model can significantly improve the accuracy of financial risk prediction and provide powerful decision support for enterprises. This research not only enriches the theoretical system in the field of financial risk management, but also provides new ideas and technical means for the practice of financial risk management and control of enterprises.

关键词： Cloud computing Analytical models Accuracy Computational modeling Knowledge graphs Predictive models Prediction algorithms data models Systems simulation Risk management

来源：评论

学校读者我要写书评

暂无评论

Detecting Malicious Domain Names from Domain Generation algorithms using Bi-directional LSTM Network

Detecting Malicious Domain Names from Domain Generation Algo...

引用

2022 international conference on Signal processing and Communication Security, ICSPCS 2022

作者： Luo, Suliang Han, Gang Li, An Peng, Jialiang School of Data Science and Technology Heilongjiang University Harbin China Heilongjiang Branch National Computer Network Emergency Response Technical Team Coordination Center of China China Institute for Cryptology and Network Security Heilongjiang University Harbin China

ISBN: (纸本)9781510660151

DNS (Domain Name System/DNS) is one of the most important infrastructures of Internet. People can easily access the rich network resources worldwide using the DNS technology. However, the Domain Generation Algorithm (DGA) is also accompanied by the DNS technology, which is used to generate malicious domain names. To detect DGA malicious domains, the previous studies often used unreal small DNS domain name datasets to train the detection models that always overlooked real user data traffic. These models generally did not have good generalization performance. In this paper, we propose a new DGA malicious domain name detection model based on Bi-directional LSTM network. We also propose a new evaluation metric to evaluate the real unlabeled DNS traffic data. Compared with LSTM model, the detection effect of our proposed model is improved effectively. The experimental results show that the precision of the model and the value of AUC reach 98.4% and 0.9079, respectively. © 2022 SPIE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

PolyDiseaseNet: A Multimodal Machine Learning Approach for Comprehensive Disease Diagnosis and Prognosis

PolyDiseaseNet: A Multimodal Machine Learning Approach for C...

引用

2023 international conference on System, Computation, Automation and Networking, ICSCAN 2023

作者： Saranya, R. Adarsh, V. Akash, S. Amirthap, P. Ram, S. Dinesh Kumar, N. Ajay Department of Artificial Intelligence and Data Science Rajalakshmi Institute of Technology Chennai India

ISBN: (纸本)9798350315127

Cancer is a leading cause of death world wide, accounting for an estimated 10 million deaths in 2020 alone. Early detection and accurate diagnosis are essential for improving patient outcomes and reducing cancer- related mortality. However, traditional diagnostic methods, such as biopsies and imaging, can be invasive, time- consuming, and expensive. Deep learning, a subfield of artificial intelligence, has emerged as a promising tool for cancer prediction and diagnosis. Deep learning algorithms can be trained on large datasets of gene expression data to identify complex patterns that are associated with diverse types of cancer. Once trained, these algorithms can be used to predict the type of cancer that a patient has based on their gene expression profile. Keras, a high- level Python library for TensorFlow, provides a user-friendly framework for building and training deep learning models. Keras offers a variety of pre-built deep learning architectures, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs). CNNs are particularly wellsuited for image analysis tasks, while RNNs are well- suited for sequence data analysis tasks. Gene expression data is a rich source of information for identifying cancer-associated patterns. Gene expression data reflects the activity of genes within cells, and different genes are expressed at distinct levels in diverse types of cells and tissues. By analyzing gene expression patterns, deep learning algorithms can identify subtle differences between cancer cells and normal cells. This study investigated the application of deep learning using Keras for cancer prediction and diagnosis, with a primary focus on five prevalent cancer types: colon, breast, lung, kidney, and prostate. The study utilized gene expression data from a large cohort of patients to develop and evaluate deep learning models. Various deep learning architectures were explored, including CNNs and RNNs. The performance of each architecture was

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

A Sensor-Based Unit for Diagnostics and Optimization of Solar Panel Installations 19th

A Sensor-Based Unit for Diagnostics and Optimization of Sola...

引用

19th European, Mediterranean, and Middle Eastern conference (EMCIS)

作者： Berntzen, Lasse Teimourzadeh, Saeed Anghelita, Paula Ming, Qian Ursu, Viorel Univ South Eastern Norway Kongsberg Norway EPRA Elect Energy Co Ankara Turkiye ICPE Bucharest Romania

ISBN: (纸本)9783031306938;9783031306945

To solve the problem of relatively low efficiency in most of nowadays solar panels, this paper proposes a new solution by diagnosing and optimizing solar panel installations. Dr. Solar, a sensor-based device, acquires information on solar radiation, geographical position, inclination, roll, panel temperature, and ambient temperature and humidity. The sensor-based device is mounted to the solar panel for a few days while collecting data. The cloud-based service gathers information from the inverter while the device is moved to other installations. Combined with the data of power generation from the inverter, a cloud service is offered for processing and analyzing the data to maximize power production by adjusting the panel installation. During several short-term experiments in different places, Dr. Solar provided reliable data for analysis on the cloud platform, and a web interface was developed to calculate power production. The solution has increased photovoltaic panels' efficiency by optimizing adjustments to the panel's physical position (tilt and angle). The paper describes the sensor-based device, the cloud service, and the algorithms used.

关键词： photovoltaic energy solar panels optimization diagnostics big data analytics sensors Dr. Solar

来源：评论

学校读者我要写书评

暂无评论

Federated Learning via Meta-Variational Dropout 37

Federated Learning via Meta-Variational Dropout

引用

37th conference on Neural Information processing Systems (NeurIPS)

作者： Jeon, Insu Hong, Minui Yun, Junhyeog Kim, Gunhee Seoul Natl Univ Seoul South Korea

ISBN: (纸本)9781713899921

Federated Learning (FL) aims to train a global inference model from remotely distributed clients, gaining popularity due to its benefit of improving data privacy. However, traditional FL often faces challenges in practical applications, including model overfitting and divergent local models due to limited and non-IID data among clients. To address these issues, we introduce a novel Bayesian meta-learning approach called meta-variational dropout (MetaVD). MetaVD learns to predict client-dependent dropout rates via a shared hypernetwork, enabling effective model personalization of FL algorithms in limited non-IID data settings. We also emphasize the posterior adaptation view of meta-learning and the posterior aggregation view of Bayesian FL via the conditional dropout posterior. We conducted extensive experiments on various sparse and non-IID FL datasets. MetaVD demonstrated excellent classification accuracy and uncertainty calibration performance, especially for out-of-distribution (OOD) clients. MetaVD compresses the local model parameters needed for each client, mitigating model overfitting and reducing communication costs. Code is available at https://***/insujeon/MetaVD.

关键词： data privacy

来源：评论

学校读者我要写书评

暂无评论

Federated Learning Integrated CNN-Trans DDoS Attack Detection Model 3

Federated Learning Integrated CNN-Trans DDoS Attack Detectio...

引用

3rd international conference on Electronic Information Engineering and Computer Science, EIECS 2023

作者： Shen, Qiaoling Wang, Yi Xie, Linzhe Zhao, Xin Bai, Junwen Zheng, Yongxin Power China Kunming Engineering Corporation Limited Kunming China

ISBN: (纸本)9798350316674

In DDoS attack monitoring [1] [2], effective detection algorithms [3] that rely on a large amount of data support are needed to achieve the desired detection results. Because individual organisations generate limited amount and characteristics of data, and considering data privacy riskiness they are reluctant to share training data [4], which often results in "data silos", and insufficient data quantity and quality can lead to inaccurate detection results. To address the above issues, this paper will explore how to use joint learning techniques to train models on massive network data while ensuring the privacy and security of organisational data. © 2023 IEEE.

关键词： data privacy data silos DDoS attack monitoring Federated learning

来源：评论

学校读者我要写书评

暂无评论

Improving Imbalanced Machine Learning with Neighborhood-Informed Synthetic Sample Placement

引用

JOURNAL OF MANAGEMENT INFORMATION SYSTEMS 2022年第4期39卷 1116-1145页

作者： Nasir, Murtaza Dag, Ali Simsek, Serhat Ivanov, Anton Oztekin, Asil Wichita State Univ W Frank Barton Sch Business Dept Finance Real Estate & Decis Sci Wichita KS USA Creighton Univ Heider Coll Business Dept Business Intelligence & Analyt Omaha NE USA Montclair State Univ Feliciano Sch Business Dept Informat Management & Business Analyt Montclair NJ USA Univ Illinois Gies Coll Business Dept Business Adm Champaign IL USA Univ Massachusetts Lowell Manning Sch Business Dept Operat & Informat Syst Lowell MA USA Pulichino Tong Business Ctr 72 Univ AveOff 442 Lowell MA 01854 USA

Machine learning is widely used in information systems design. Yet, training algorithms on imbalanced datasets may severely affect performance on unseen data. For example, in some cases in healthcare, fintech, or cybersecurity contexts, certain subclasses are difficult to learn because they are underrepresented in training data. Our study offers a flexible and efficient solution based on a new synthetic average neighborhood sampling algorithm (SANSA), which, in contrast to other solutions, introduces a novel "placement " parameter that can be tuned to adapt to each dataset's unique manifestation of the imbalance. This package can be downloaded for R- 1 . We tested SANSA against seven existing sampling methods used in conjunction with the four most frequently used machine learning models trained on 14 benchmark datasets. Our results provide suggestive evidence that SANSA offers a feasible solution to the imbalance problem for most datasets. Our findings provide practical recommendations for how SANSA can be effectively implemented while reducing the complexity level of an imbalanced learning pipeline.

关键词： Imbalanced data oversampling undersampling machine learning predictive analytics classification prediction performance algorithm training

来源：评论

学校读者我要写书评

暂无评论

Automatic Diagnosis of Age-Related Macular Degeneration via Federated Learning 8th

Automatic Diagnosis of Age-Related Macular Degeneration via ...

引用

8th international conference on Computer Vision and Image processing (CVIP)

作者： Sonti, Movya Kokil, Priyanka Indian Inst Informat Technol Design & Mfg Dept Elect & Commun Engn Adv Signal & Image Proc ASIP Lab Chennai 600127 Tamil Nadu India

ISBN: (纸本)9783031581731;9783031581748

Artificial intelligence has taken healthcare a step forward by providing quick diagnosis and treatment recommendations. The traditional machine-learning approach requires massive data to train the model for better diagnosis. But, due to policy regulations, medical data is always guarded by the barricades of the law, making data accessibility difficult for researchers. To address this issue, a data-decentralized collaborative framework known as federated learning is adopted that reaps the benefits of huge private data without aggregating it into a single common store. A pre-trained model is employed to diagnose age-related macular degeneration and performance of the proposed framework is compared with other two model architectures, namely, MobileNet and InceptionV3. To investigate the effectiveness of the proposed framework, a comparison is made with a data-centralized learning approach. Using the MobileNet model, the federated and centralized frameworks have achieved an accuracy of 95% and 92%, respectively. These findings encourage clinicians around the globe to utilize wealthy private data without violating privacy laws using federated learning to build a powerful model for classifying any disorders while maintaining data privacy.

关键词： Federated learning Privacy data-security data-decentralization Pre-trained models Age-related macular degeneration

来源：评论

学校读者我要写书评

暂无评论

22nd international conference on Big data Analytics and Knowledge Discovery, DaWaK 2020

22nd International Conference on Big Data Analytics and Know...

引用

22nd international conference on Big data Analytics and Knowledge Discovery, DaWaK 2020

ISBN: (纸本)9783030590642

The proceedings contain 30 papers. The special focus in this conference is on Big data Analytics and Knowledge Discovery. The topics include: Extreme-sax: Extreme points based symbolic representation for time series classification;framework to optimize data processing pipelines using performance metrics;a scalable randomized algorithm for triangle enumeration on graphs based on sql queries;data engineering for data science: Two sides of the same coin;mining attribute evolution rules in dynamic attributed graphs;sustainable development goal relational modelling: Introducing the sdg-cap methodology;mining frequent seasonal gradual patterns;derivative, regression and time series analysis in sars-cov-2;building a competitive associative classifier;contrastive explanations for a deep learning model on time-series data;Dhe $$^models$$ : Distributed Hybrid Evolution Engine for Performance Optimizations of Computationally Intensive Applications;cyberbullying detection in social networks using deep learning based models;predicting customer churn for insurance data;scalable machine learning on popular analytic languages with parallel data summarization;which bills are lobbied? predicting and interpreting lobbying activity in the us;fibs: A generic framework for classifying interval-based temporal sequences;multivariate time series classification: A relational way;behave or be detected! identifying outlier sequences by their group cohesion;detecting anomalies in production quality data using a method based on the chi-square test statistic;learning from past observations: Meta-learning for efficient clustering analyses;parallel k-prototypes clustering with high efficiency and accuracy;grand reports: A tool for generalizing association rule mining to numeric target values;self-organizing map for multi-view text clustering;expected vs. unexpected: Selecting right measures of interestingness.

关键词：

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：