检索结果-内蒙古大学图书馆

12th international Advanced Computing conference, IACC 2022

作者： Rizvi, Fizza Mohi ud din, Saika Sharma, Nonita Sharma, Deepak Kumar Department of Information Technology Indira Gandhi Delhi Technical University for Women Kashmere Gate Delhi India

ISBN: (纸本)9783031356438

Currently, there are millions and trillions of webpages online. Therefore, in order to protect their data, users must be able to distinguish between trustworthy and dangerous websites. Since they are on the rise, malicious URLs (Uniform Resource Locators) are one of the most common cybersecurity concerns nowadays. They host free content (like malware, ransomware, trojans, password sniffers, etc.), and by clicking on this free content, especially the links found in emails, messages, and other formats, unwary users become the targets of these scams, have their personal information compromised, or have their system infected with malware, resulting in the loss of billions of rupees. There are more threats to cybersecurity than ever before, so a quick fix is required. In this study, we investigate how well different machine learning methods work in spotting potentially dangerous URLs. To identify both dangerous and benign URLs, this study employs a variety of machine learning methods. Decision Tree classifier, Random Forest classifier, AdaBoost classifier, K-Nearest Neighbors classifier, Gaussian Nave Bayes classifier, and XGB classifier are just a few of the machine learning methods that are used. Finding the most effective algorithm for accurately detecting fraudulent URLs is the main objective of this project. The accuracy metric was utilised as the evaluation tool to determine the accuracy of the various algorithms, with Random Forest having the highest accuracy (92.6%) of all of them. © 2023, Springer Nature Switzerland AG.

关键词： Websites

来源：评论

学校读者我要写书评

暂无评论

Early Alzheimer Detection Through Speech Analysis and Vision Transformer Approach 1st

Early Alzheimer Detection Through Speech Analysis and Visio...

引用

1st international conference on Speech and Language Technologies for Low-resource Languages, SPELLL 2022

作者： Pranav, G. Varsha, K. Gayathri, K.S. Information Technology Sri Sivasubramaniya Nadar College of Engineering affiliated to Anna University Kalavakkam Chennai603110 India

ISBN: (纸本)9783031332302

Alzheimer’s disease is a brain disorder which slowly destroys and deteriorates memory and thinking abilities and, eventually, the ability to carry out the basic tasks. Early diagnostic of Alzheimer’s’s disease can avail the patient with high probability of cure with minimal cost. The cost required for detection of Alzheimer’s is quite high for common people. Therefore diagnosation of Alzheimer from spontaneous speech proves to be a feasible and convenient method as it can be done at home. This paper implements Log Mel Spectrograms with vision trans- former deep learning classifier approach to differentiate the Alzheimer group and control group. The vision transformer uses a transformer like architecture over the patches of the image in order to classify the images. Extensive analysis done over the Pitt corpus dataset with the help of the proposed approach, demonstrates the efficiency of Vision Transformer and log mel spectrogram approach over the MFCC and random forest classifier approach. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

A Survey of Statistical, machine learning, and Deep learning-Based Anomaly Detection Techniques for Time Series 14th

A Survey of Statistical, Machine Learning, and Deep Learning...

引用

12th international Advanced Computing conference, IACC 2022

作者： Oswal, Sangeeta Shinde, Subhash Vijayalakshmi, M. AI and Data Science VESIT Mumbai India Computer Engineering LTCE Navi Mumbai India

ISBN: (纸本)9783031356438

Anomaly detection, also known as outlier detection, has been a persistent yet active subject of study in a number of research areas for several decades. This study provides a complete taxonomy of the research on anomaly detection in time series related to types of time series data, types of anomalies, detection methods, and the dataset of time series. This taxonomy simplifies the understanding of the procedures related to each group. In addition, we detail the benefits and drawbacks of each type of approach under each respective category. Recent work based on deep learning has made significant strides in this sector. They are capable of discovering data abnormalities and learning unsupervised representations of large-scaled sequences. This survey aims to facilitate a better understanding of the many research approaches taken on this topic since the research community lacks a comprehensive comparison of statistics, machine learning, and deep learning techniques. Despite substantial progress in this area of research, there is no one anomaly detector that has been demonstrated to be effective across several datasets. Current anomaly detection methods struggle to detect anomalies related to system actions or context. © 2023, Springer Nature Switzerland AG.

关键词： Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

Computer-Aided Design of Electric Drives with FEA software 41st

Computer-Aided Design of Electric Drives with FEA Software

引用

41st IBIMA international conference on Artificial intelligence and machine learning, IBIMA-AI 2023

作者： Manka, Michal Karpiel, Grzegorz Prusak, Daniel RIOT Sp. z o.o. ul. M. i B. Wyslouchów 26/26 30-612 Kraków Poland Faculty of Mechanical Engineering and Robotics Department of Robotics and Mechatronics AGH University of Krakow Al. Mickiewicza 30 Kraków30-059 Poland

ISBN: (纸本)9783031628429

The rapid development of new technologies in recent years leads to a significant reduction in time-to-market. It requires a significant increase in research and development expenditure or a change in the design process. Such a change was the introduction of virtual prototyping techniques, i.e.: Finite Element Analysis (FEA). Initially, FEA was limited to mechanics, however, the introduction of multi-domain simulation, allows to study other physical phenomena, such as thermal or electromagnetic, which are of interest to us in this case. Often in commercial applications, time pressure limits the scope of research for an optimal solution to shorten the research time, which leads to suboptimal solutions. In the case presented in the paper, virtual prototyping techniques were used in the design process of the BLDC direct drive using FEA techniques, the main criterion of which was to shorten the design time while maintaining a wide range of tests and high test accuracy. The proposed methodology assumes using 2D simulations for approximate tests and then accurate 3D simulations only for selected, most promising configurations. Approximate 2D tests are several times faster than 3D tests, which means that it can be tested significantly more configurations at the same time, which increases the chance of finding the global optimum of the structure. The use of FEA 3D tests for selected constructions significantly improves the accuracy of the results obtained and allows for the final selection of the best solution. The example presented in the paper allowed for the design of the BLDC motor, taking into account the testing of 31 different BLDC drive configurations, and thus determining its optimal design, confirmed by the construction of the final prototype. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Finite element method

来源：评论

学校读者我要写书评

暂无评论

Video Tampering Detection Using machine learning and Deep learning 14th

Video Tampering Detection Using Machine Learning and Deep Le...

引用

12th international Advanced Computing conference, IACC 2022

作者： Deo, Siddhi Mehta, Simran Jain, Digha Tiwari, Charu Thorat, Aniket Mahara, Sudhanshu Gonge, Sudhanshu Joshi, Rahul Gite, Shilpa Kotecha, Ketan Department of Computer Science and Engineering Symbiosis Institute of Technology Pune India

ISBN: (纸本)9783031356438

Digital recordings are now affordable, quick to make, and straightforward to upload on web platforms because of the widespread availability of video recording technology in smartphones and other digital devices. The truth of films published on social media cannot be trusted. With the latest tools used for video editing tools, movies may now be readily changed (faked) for political ends or unlawful profit. Over the years, a lot of effort has gone into creating novel methods for identifying various forms of video manipulation. The objectives of video tampering detection include identifying signs of change and assessing the validity and integrity of the video file. These methods are divided into two groups: active & passive (blind) techniques. Based on the forgeries that passive video tampering detection systems can see, they may be divided into three groups: region tampering detection, double or multiple compressed video detection, and video inter-frame forgery detection. The previous studies under consideration are provided in a condensed tabular format, together with the criteria they employed and any limitations they had. The research work concluded by outlining a number of problems that might be used to inspire new research topics for the identification of passive video manipulation. In this paper, few AI/ML algorithms were tested for video tampering detection to find out which one gives out the most accurate results. © 2023, Springer Nature Switzerland AG.

关键词： Digital devices

来源：评论

学校读者我要写书评

暂无评论

Preface

引用

Journal of Physics: conference Series 2023年第1期2575卷 011001页

作者： Yu, Yury Asia Pacific Institute of Science and Engineering

2023 7th international conference on Data Mining, Communications and Information Technology (DMCIT 2023), was successfully held on May 27, 2023, which was organized by Asia Pacific Institute of Science and engineering (APISE), and sponsored by Jiangnan University. DMCIT conference has been held for six years since 2017 after successful onsite conferences or virtual conferences. DMCIT 2023 continued its tradition of providing a unique and excellent forum for exchanging research methodologies, exploring practical applications, and fostering innovative *** 2023 strives to offer networking opportunities to meet and interact with the world-leading scientists, engineers and researchers as well as industrial partners in Core Data Mining, Communication engineering, Information Technology Applications *** conference was originally scheduled in Chongqing, China from May 26 to 28, 2023. Most authors cannot attend offline with personal reasons, but they can only attend online. As a result of small number of offline participants, it had to be turned into a fully virtual conference via Tencent meeting on May 27, *** proceedings of this year’s edition comprised four main categories: “1. Image Processing and Computer Vision”; “2. machine learning and Data Analysis”; “3. Information Technology and Communication Systems”; “4. Optimization and Modeling”. All these submissions were rigorously reviewed by the Program Committee. The research tracks attracted 42 submissions in total, and out of 27 papers were accepted, including countries like Pakistan, Japan, Philippines and China, *** the conference day, 9 oral presentations and 18 posters presentation were arranged according to the participants’ choices. Two awards, one best oral presentation award and one best poster presentation award were selected by the end of conference. Four distinguished keynote speakers, including Prof. Shahram Latifi from University of Nevada, Las Vegas (USA); Prof. Yudong Zhang from The

关键词：

来源：评论

学校读者我要写书评

暂无评论

Self Supervised learning for Classifying the Rotated Images 14th

Self Supervised Learning for Classifying the Rotated Image...

引用

12th international Advanced Computing conference, IACC 2022

作者： Roy, Sidharth Marathe, Aboli Walambe, Rahee Kotecha, Ketan Maharashtra Pune412 115 India Maharashtra Pune412 115 India Maharashtra Paud412 108 India Machine Learning Department Carnegie Mellon University PittsburghPA15213 United States

ISBN: (纸本)9783031356438

As the applications of image classification models have grown, there has emerged a need to develop classifier models which are invariant to the orientation of the image. Traditionally, image augmentation is applied during training of classification models to encourage the quality of rotation invariance. In our work, we propose an alternate method of classifying rotated images in which the 2d rotation transformation applied to an image is identified and reverted before being passed to an image classifier, thus increasing classification accuracy. The rotation correction is achieved by feeding the input images through a Rotnet, a deep convolutional model which has been trained to identify the orientation of images, and rotating the images back to the original orientation. We find that our method achieves an increase in image classification accuracy over the baseline model on the CIFAR10 dataset. Specifically, when images are rotated and passed to an image classifier with an accuracy of 94% on the CIFAR10 classification task, the accuracy drops to 59.1%. However when the randomly rotated input images were corrected for rotation using our method before being fed to the classifier, the accuracy increases to 75.8%. © 2023, Springer Nature Switzerland AG.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

DR-A-LSTM: A Recurrent Neural Network with a Dimension Reduction Autoencoder a Deep learning Approach for Landslide Movements Prediction 14th

DR-A-LSTM: A Recurrent Neural Network with a Dimension Reduc...

引用

12th international Advanced Computing conference, IACC 2022

作者： Kumar, Praveen Priyanka Uday, K.V. Dutt, Varun ACS Lab Indian Institute of Technology Mandi Himachal Pradesh Mandi175075 India Geotechnical Engineering Lab Indian Institute of Technology Mandi Himachal Pradesh Mandi175075 India

ISBN: (纸本)9783031356438

Landslides are a challenging problem in India and the world. Different weather conditions and soil properties could trigger landslides. The machine learning (ML) models could predict landslides’ movements. The ML model may overfit the high-dimensional feature of weather and soil data. Dimension reduction techniques could reduce the dimension of the features. The autoencoder model was developed to reduce the data dimension in this experiment. The four months, April to August 2022, time series data of the landslide monitoring station (LMS) from the five landslides in Himachal Pradesh, India, were considered to train the ML models. The dimension reduction autoencoder long-short-term memory (DR-A-LSTM), an autoencoder and LSTM model ensemble, was developed. Furthermore, the ensemble of principal component analysis (PCA) and LSTM (PCA-LSTM) model was developed to reduce the dimension by PCA. The DR-A-LSTM model was compared with the simple LSTM and PCA-LSTM models to predict landslide movements. The data was split in the 80:20 ratio to train and test the ML models. The simple LSTM model produced 82.3% accuracy in the training data and 71.8% in the testing data. The simple LSTM model showed overfitting in the training data. The PCA-LSTM model produced 76.5% accuracy in training and 88.2% in testing. Next, the DR-A-LSTM model produced 97.8% accuracy in training and 100% in testing. The findings of this experiment suggest that the DR-A-LSTM model performed better than the simple LSTM model and PCA-LSTM. As a result, the DR-A-LSTM model could be developed for real-time landslide movement predictions. © 2023, Springer Nature Switzerland AG.

关键词： Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

Univariate, Multivariate, and Ensemble of Multilayer Perceptron Models for Landslide Movement Prediction: A Case Study of Mandi 14th

Univariate, Multivariate, and Ensemble of Multilayer Percept...

引用

12th international Advanced Computing conference, IACC 2022

作者： Priyanka Kumar, Praveen Devi, Arti Akshay, K. Gaurav, G. Uday, K.V. Dutt, Varun ACS Lab Indian Institute of Technology Mandi Himachal Pradesh Mandi175075 India Geotechnical Engineering Lab Indian Institute of Technology Mandi Himachal Pradesh Mandi175075 India

ISBN: (纸本)9783031356438

The landslides are a challenging problem in the Himalayan states such as Himachal Pradesh and Uttarakhand in India and as well as in the world. machine learning models could be developed to predict the movement of landslides in advance. In our proposed study, we developed a univariate, multivariate, and ensemble multilayer perceptron (MLP) and trained on the data collected from ten different stations in the Mandi district in Himachal Pradesh, India. The primary goal of this paper is to develop a model to predict the movement value using time series data. Recorded data was divided into the 80:20 ratio to train and test the model’s performance. The root-mean-squared error (RMSE) was used to compare the model’s performance. In training, the multivariate MLP was the best model with 0.012 RMSE, and ensemble MLP was the second-best model with 0.025 RMSE. In testing, the ensemble MLP was the best with 0.012 RMSE, and the univariate MLP was the second-best model with 0.013 RMSE. The analysis of the results shows that ensemble MLP is a promising method that can be used for landslide prediction using movement data. © 2023, Springer Nature Switzerland AG.

关键词： Landslides

来源：评论

学校读者我要写书评

暂无评论

A Chi-Square Dissimilarity Measure for Clustering Categorical Datasets 14th

A Chi-Square Dissimilarity Measure for Clustering Categoric...

引用

12th international Advanced Computing conference, IACC 2022

作者： Cardona, Luis Ariosto Serna Hernández, Kevin Alejandro Barreto, Gina Marcela González, Piedad Navarro Gutiérrez, Álvaro Ángel Orozco Department of Electrical Engineering Universidad Tecnológica de Pereira Pereira Colombia Pereira Colombia

ISBN: (纸本)9783031356438

Currently, there has been a high use of databases of large proportionalities. In addition, in using these data, there has been an enormous increase in using categorical data, specifically in using new alternatives to identify the most relevant items. In this order of ideas, cluster analysis is a relevant approach for the processing of categorical data. However, different machine learning models that have been proposed in the literature have problems to interpret categorical variables because of their high dimensionality and data overlapping, which can cause high computational cost or low performance in the algorithms. For this reason, we propose an unsupervised method using the C-S (Chi-Square) dissimilarity measure mapping from a categorical to a continuous Euclidean space, allowing an adequate interpretation of the k-means algorithm with the squared Euclidean distance. Furthermore, the proposed method was compared with other state-of-the-art techniques in unsupervised learning for categorical data such as: k-means, Mkm-nof, weighted dissimilarity, Mkm-ndm and structure-based clustering (SBC) algorithms;evaluated the accuracy (AC), adjusted rand index (ARI) and normalized mutual information (NMI). The results we present in the proposal outperform the clustering methods in the different evaluation methods on the 9 databases worked, for example, the (AC) of our Kmeans (C-S) method presented on the whole dataset is 0. 8090, in SBC1 of 0.7907, SBC2 of 0.7820, k-modes of 0.6979, W-D of 0.6949, Mkm-not of 0.6906 and Mkm-ndm of 0.7254 demonstrating superiority not only in the AC, but also in NMI and ARI. On the other hand, computational time was an issue of great relevance in our proposal because of the results got, in them it can be interpreted that the Kmeans (C-S) method in 8 of the 9 databases takes less than half the time of the other algorithms executing its model. © 2023, Springer Nature Switzerland AG.

关键词： Database systems

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：