检索结果-内蒙古大学图书馆

14th International Conference on Computing Communication and Networking Technologies, ICCCNT 2023

作者： Nikhil, Shreyas Sah, Rahul Kumar Kumar Parki, Santosh Tamang, Til Bikram Somashekhara Reddy, D. Mahesh, T.R. Faculty of Engineering and Technology Dept. of Computer Science and Engineering Bangalore India

ISBN: (纸本)9798350335095

Stock price prediction has always been a tough task for all the stakeholders involved. This paper focusses on four different models, namely LSTM, CNN, LSTM-CNN, and Genetic Algorithm-Assisted LSTM-CNN (GA-LSTM-CNN) for stock price prediction. The day-to-day stock prices of a certain company for a period of more than 10 years make up the dataset utilized in this study. Root mean square error (RMSE) has been utilized as performance metric for each model. Outcomes of our experiment show that the GA-LSTM-CNN model performed better than other models with an RMSE value of 0.0064 upon training and an RMSE value of 0.03 upon testing. The GA-LSTM-CNN model combines the strengths of LSTM as well as CNN models with the help of a genetic algorithm, which helps in optimizing the hyperparameters. According to the results, the GA-LSTM-CNN model can be very helpful for the stakeholders in predicting future stock values. © 2023 IEEE.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Balancing Safety and Security in Autonomous Driving Systems: A Machine Learning Approach with Safety-First Prioritization 40

Balancing Safety and Security in Autonomous Driving Systems:...

引用

40th Annual computer Security applications Conference Workshops, ACSACW 2024

作者： Hasani, Afshin Nia, Mehran Alidoost Atani, Reza Ebrahimi University of Guilan Dept. of Computer Engineering Rasht Iran Shahid Beheshti University Faculty of Computer Science and Engineering Tehran Iran

ISBN: (纸本)9798331532819

In the era of autonomous vehicles (AVs), ensuring the safety and security of the system is an ongoing challenge, particularly when faced with increasingly cyber-Attacks such as GPS spoofing and man-in-The-middle. This paper presents a novel reinforcement learning (RL)-based framework that prioritizes safety over security concerns in AVs, ensuring that life-critical tasks such as collision avoidance and speed control are always maintained, even in the presence of security threats. Our system leverages the MAPE-K loop to dynamically adapt to changing conditions, using RL in the Analyzing phase to make real-Time decisions that maximize cumulative safety rewards while addressing security risks that directly affect operational integrity. We demonstrate how our RL-based self-Adaptive system mitigates security threats while maintaining safe driving behaviors through various simulations and experiments. The proposed system shows significant improvements in collision avoidance and threat detection, ultimately ensuring that the AV's safety is preserved in even the most challenging scenarios. © 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

VAR: An Efficient Silent Video to Speech System with Subtitle Generation using Visual Audio Recall 5

VAR: An Efficient Silent Video to Speech System with Subtitl...

引用

5th International Conference on Inventive Research in Computing applications, ICIRCA 2023

作者： Shanmugam, Diya Dhandapani Syed, Safiya Fathima Dinesh, Sahana Chitrakala, S. College of Engineering Guindy Dept. of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350321425

In this paper, a method to generate speech from a muted video is proposed. Recent advances in computer vision and deep learning have allowed the development of lip-reading technology that can recognize and interpret lip movements in videos. However, one of the main challenges in visual-to-speech synthesis is distinguishing between homophenes, or sets of words that look identical on a person's lips but contain different sounds. To address this challenge, the proposed model utilizes a speech generation framework that uses an external memory called Visual Audio Recall (VA-Recall) to generate speech from silent video. The VA-Recall system allows the system to effectively synthesize speech from a speaker by recalling the most relevant auditory information that matches the visual input context. The proposed method further generates subtitles for the reconstructed audio. Overall, this process allows for a more natural and contextually appropriate speech synthesis system. © 2023 IEEE.

关键词： Speech synthesis

来源：评论

学校读者我要写书评

暂无评论

Optimal Scheduling in Multiserver Queues

引用

Performance Evaluation Review 2024年第3期51卷 29-32页

作者： Grosof, Isaac Carnegie Mellon University Computer Science Department United States Northwestern University Dept. of Industrial Engineering & Management Sciences United States

Scheduling theory is a key tool for reducing latency (i.e. response time) in queueing systems. Scheduling, i.e. choosing the order in which to serve jobs, can reduce response time by an order of magnitude with no additional resources. Scheduling theory is well-developed in single-server systems, where one job is processed at a time. However, little is known about scheduling in multiserver systems, where many jobs are processed at once. Results are especially limited in stochastic multiserver scheduling theory. Today’s data-centers have thousands of servers, and scheduling theory is unable to analyze such systems. My thesis proves the first optimality results and first closed-form bounds on mean response time for scheduling policies in stochastic multiserver models which reflect the behavior of modern computing systems. My thesis proves novel optimality results in each of three areas: I start by studying one-server-per-job multiserver models, and prove the first results on optimal scheduling in that setting. Next, I study the multiserver-job (MSJ) model, where different jobs require different amounts of resources to be served. I prove the first characterization of mean response time for any scheduling policy in the MSJ model, as well as the first optimality results. Finally, I study the effects of scheduling on the tail of response time, rather than mean response time. I invent a novel scheduling policy, Nudge, which I prove to be the first policy to outperform FCFS’s asymptotic tail of response time. This extended abstract briefly outlines the main contributions of my thesis, describing my novel performance bounds and optimality results in each of these three areas. © 2024 Association for Computing Machinery. All rights reserved.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Comparative Study on Extractive Summarization Using Sentence Ranking Algorithm and Text Ranking Algorithm

Comparative Study on Extractive Summarization Using Sentence...

引用

2023 International Conference on Power, Instrumentation, Control and Computing, PICC 2023

作者： Majeed, Mansoora Kala, M.T. Government Engineering College Dept. of Computer Science and Engineering Thrissur India

ISBN: (纸本)9798350334463

It is possible to create an accurate and succinct summary using a method called automatic text summarising. Before creating the necessary summary sentences, the machine learning algorithms may be trained to understand documents and recognise the portions that transmit crucial facts and information. Abstractive text summa- rization and Extractive text summarization are its two methods. An extractive text summarization refers to the process of removing significant text from a text document or source manuscript. Deep learning techniques are used to train an abstractive text summarizer, which then generates new phrases and keywords based on the content of the original text. The process of acquiring and assimilation of the knowledge from numerous sources takes time. Automatic text summary is used to speed up the process of finding and consuming pertinent information. An extractive text summarising technique chooses crucial phrases, sentences, and other parts of the original text and concatenates them into a compressed form. The extractive text summarization produces grammatically correct sentences. The project is to develop a model that demonstrates extractive text summarization using sentence ranking and text rank algorithm. It shows that the text rank algorithm does a remarkable performance to extractive text summarization. Using a sentence ranking algorithm, extractive summaries are created that rate sentences according to their weights. In order to create a high-quality summary of the input document and store the summary as audio, the input document's highly scored sentences are extracted. A non-supervised, extractive text summarising method is called TextRank. While calculating the value of words and phrases in a document, the TextRank algorithm looks at how they relate to one another. According on how important the words are in each phrase, it then assigns a score to each one. The summary is formed using the algorithm's choice of the most crucial phrases. © 2023 IE

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Smart City Surveillance: Survey on Current Status and Future Trends 14

Smart City Surveillance: Survey on Current Status and Future...

引用

14th International Conference on Advances in Computing, Control, and Telecommunication Technologies, ACT 2023

作者： Neyaz, Hera Paul, Hrituparna Srivastava, Anima Singh, Shiksha Dept. of Computer Science and Engineering Shambhunath Institute of Engineering and Technology India

This paper presents a comprehensive survey on surveillance and security aspects of a smart city. Multiple dimensions of a smart city surveillance system have been discussed. The current trends and methodologies are thoroughly researched and presented. The algorithms of novel methods are reviewed along with analytics and methodologies. Impact of XAI in the smart city surveillance is discussed. Advantages and disadvantages of current infrastructures have been included in the broad literature survey. Lastly, the challenges and future trends are conferred to. © Grenze Scientific Society, 2023.

关键词： Smart city

来源：评论

学校读者我要写书评

暂无评论

DeepStock Forecast: Unveiling Market Movements through Advanced Deep Learning Models 4

DeepStock Forecast: Unveiling Market Movements through Advan...

引用

4th International Conference on Emerging Systems and Intelligent Computing, ESIC 2024

作者： Das, Smruti Rekha Mishra, Debahuti Lenka, Arundhati Shaw, Kailash Gitam Deemed to Be University Dept. Computer Science and Engineering Andhra Pradesh Visakhapatnam India Siksha 'o' Anusandhan Deemed to Be University Odisha Dept. Computer Science and Engineering Odisha India Dept. of Aiml Symbiosis Institute of Technology Pune India

ISBN: (纸本)9798350349856

This study uses deep neural network (DNN) methodologies, including auto encoder, deep belief network (DBN), and backpropagation neural network (BPNN), to predict stock prices over 15 and 30 days. Utilizing BSE Sensex and NSE Sensex datasets with diverse technical indicators, the research highlights DNN's superior training accuracy, measured by minimal mean squared error (MSE). The study's key contribution is applying DNN to predict stock market volatility. While acknowledging the limitation of relying solely on training error, the study incorporates additional measures like RMSE, MAPE, MAE, and ARV during testing. Notably, DNN's MAPE results fall within a narrow range (0.0221 to 0.0255), indicating minimal error deviation. In a comprehensive evaluation across all datasets, DNN consistently outperforms DBN and BPNN in both training and testing performance. © 2024 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

A Deep Learning Based Image Processing Technique for Early Lung Cancer Prediction

A Deep Learning Based Image Processing Technique for Early L...

引用

2024 ASU International Conference in Emerging Technologies for Sustainability and Intelligent Systems, ICETSIS 2024

作者： Tasnim, Nowshin Noor, Kazi Rifah Islam, Mursalina Huda, Mohammad Nurul Sarker, Iqbal H. Chittagong University of Engineering & Technology Dept. of Computer Science & Engineering Chittagong4349 Bangladesh United International University Dept. of Computer Science & Engineering Dhaka1212 Bangladesh Edith Cowan University PerthWA6027 Australia

ISBN: (纸本)9798350372229

Lung cancer is the primary cause of cancer mor-tality all over the world due to the increase of tobacco consumption, and industrialization in developing nations. As the early-stage diagnosis can reduce the mortality rate significantly, early detection with the availability of high-tech Medical facilities is highly necessary. In this research, we used deep learning (DL) methods initially on patient's 1190 CT scan images from the Kaggle IQ-OTH lung cancer dataset, and after significant image preprocessing steps we found augmented images including normal, malignant, and benign cases to identify high-risk in-dividuals to detect lung cancer and also predict the malignancy and thus, taking early actions to prevent long-term consequences. A thorough performance comparison between several classifiers, including the conventional CNN, Resnet50, and InceptionV3, has been presented. Here, affine transformation, gaussian noise, and other rigorous image preprocessing techniques are used. The contribution obtained a 98% validation accuracy while reducing the model's complexity with the previous preprocessing stage. The comparison method shows that the suggested preprocessing method yields a higher F1 score value of 97%, validating our suggested methodology. © 2024 IEEE.

关键词： CT scans deep learning healthcare analytics image processing lung cancer prediction

来源：评论

学校读者我要写书评

暂无评论

A Critical Analysis of Object Tracking Using Radial Function Networks

A Critical Analysis of Object Tracking Using Radial Function...

引用

2024 International Conference on Smart Devices, ICSD 2024

作者： Mishra, Amit Kumar Pandey, Neeraj Kumar Diwakar, Manoj Bahadure, Nilesh Bhaskarrao Dumka, Ankur Patni, Jagdish Chandra Graphic Era Hill University Dept. of Computer Science & Engineering Dehradun India Graphic Era Deemed to be University Dept. of Computer Science & Engineering Dehradun India Symbiosis Institute of Technology Dept. of CSE Nagpur Campus Pune India Women Institute of Technology Dept. of Computer Science & Engineering Dehradun India Alliance University Alliance School of Advance Computing Dept. of CSE Karnataka Bengaluru India

ISBN: (纸本)9798350394412

A important method in several fields, including driverless cars, medical imaging, human-computer interaction, and investigation systems, is object tracking. Because of their flexibility to changes in object presence and ambient factors, as well as their capacity to estimated complex, non-linear functions, radial networks have become a potent tool for object stalking in recent years. The article conditions for the analysis of the radical working networking where the algorithm is a tracking segment that aims in moving the object. Object recognition is one of the most basic and centralized factor analyses in determining outputs having relevant values and a strong research background. Radial functioning is a phenomenon that has gained rapid training and desirable properties for improving the presentation of networking sites. radial networks, a form of artificial neural network, utilize radial basis functions as initiation functions. Because of this characteristic, RBF networks are especially well-suited for applications like object tracking that call for high accuracy in function approximation and pattern recognition. We explore the design of RBF networks in detail, emphasizing the functions of the input layer, output layer, and hidden layer using radial basis functions. We examine the mathematical foundations that allow RBF networks to represent intricate connections in the data, emphasizing how well they can generalize and interpolate. © 2024 IEEE.

关键词： Radial basis function networks

来源：评论

学校读者我要写书评

暂无评论

Distributionally Robust Clustered Federated Learning: A Case Study in Healthcare 63

Distributionally Robust Clustered Federated Learning: A Case...

引用

63rd IEEE Conference on Decision and Control, CDC 2024

作者： Konti, Xenia Riess, Hans Giannopoulos, Manos Shen, Yi Pencina, Michael J. Economou-Zavlanos, Nicoleta J Zavlanos, Michael M. Duke University School of Medicine United States Duke University Dept. of Computer Science United States Duke University Dept. of Electrical and Computer Engineering United States Duke University Dept. of Mechanical Engineering & Materials Science United States

ISBN: (纸本)9798350316339

In this paper, we address the challenge of heterogeneous data distributions in cross-silo federated learning by introducing a novel algorithm, which we term Cross-silo Robust Clustered Federated Learning (CS-RCFL). Our approach leverages the Wasserstein distance to construct ambiguity sets around each client's empirical distribution that capture possible distribution shifts in the local data, enabling evaluation of worst-case model performance. We then propose a model-agnostic integer fractional program to determine the optimal distributionally robust clustering of clients into coalitions so that possible biases in the local models caused by statistically heterogeneous client datasets are avoided, and analyze our method for linear and logistic regression models. Finally, we discuss a federated learning protocol that ensures the privacy of client distributions, a critical consideration, for instance, when clients are healthcare institutions. We evaluate our algorithm on synthetic and real-world healthcare data. © 2024 IEEE.

关键词： Federated 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：