检索结果-内蒙古大学图书馆

2025 International Conference on Pervasive Computational Technologies, ICPCT 2025

作者： Akkalkot, Afsha Imran Kulshrestha, Nitin Sharma, Geeti Singh Sidhu, Kawerinder Palimkar, Sneha S. Nethravathi, K. College of Engineering Pune Technological University Computer Science Engineering Maharashtra Pune India Christ Deemed to Be University Commerce Finance and Accountancy Uttar Pradesh Ghaziabad India Bengaluru India Uttaranchal Institute of Management Uttaranchal University Uttarakhand Dehradun India Computer Engineering and Technology Coep Technological University Maharashtra Pune India University Bangalore India

ISBN: (纸本)9798331508685

Artificial intelligence (AI) has been used more and more in financial decision-making recently, raising questions about the accountability and transparency of these complex systems. The current study investigates the way Explained Artificial Intelligence (XAI) methods might alleviate these concerns and improve the openness of financial decision-making procedures. Nowadays machine learning algorithms are easier to use than ever before, but creating and deploying systems that facilitate real-world banking services has proved challenging. This is mostly due to the fact that algorithms for machine learning are neither transparent or explainable, two attributes that are essential to creating reliable technology. What sets this study unique is the construction of an explainable artificial intelligence (XAI) model that addresses these accessibility concerns while also serving as an instrument for the establishment of credit risk control policies. This work proposes an explainable artificial intelligence model for financing risk control to measure the risks associated with credit financing via peer-to-peer financing networks. The framework uses Shapley parameters to provide AI forecasts according to significant factors that explain. The Support Vector Machine (SVM) and gradient boosting methods had the greatest accuracy scores, 92.4 and 97.6, accordingly. The accuracy of the model was evaluated on a bigger database, and the findings demonstrated that it regularly achieved high levels of accuracy. The SVM and GBM models achieved accuracies of 94.8 and 97.6, respectively. © 2025 IEEE.

关键词： Decentralized finance

来源：评论

学校读者我要写书评

暂无评论

AI-Driven Incident Response Systems for Cybersecurity: A Hybrid Approach 12th

AI-Driven Incident Response Systems for Cybersecurity: A Hyb...

引用

12th International Conference on Frontiers of Intelligent Computing: Theory and Applications, FICTA 2024

作者： Soni, Monika Pawar, Avinash M. Godbole, Aditee Sharma, Ashutosh Tiwaskar, S.A. Kokane, Chandrakant Deelip Auricle Global Society of Education and Research Bikaner India Bharati Vidyapeeth’s College of Engineering for Women Maharashtra Pune India Nagpur Campus Pune India Electronics and Communication Engineering Arya College of Engineering Jaipur India Computer Engineering VIIT Pune India Nutan Maharashtra Institute of Engineering and Technology Maharashtra Pune India

ISBN: (纸本)9789819601424

The increasing complexity and sophistication of cybersecurity threats necessitate the development of strong and flexible incident response systems. This study presents a hybrid methodology that utilizes various machine learning models such as Random Forest (RF), Decision Trees (DT), Isolation Forest, Convolutional Neural Network (CNN), and a unique Hybrid Gated Recurrent Unit—CNN (GRU-CNN) design. The proposed model seeks to improve the precision and effectiveness of incident detection and response in real-time. The study shows that combining various models is effective in creating a robust defense mechanism against cyber threats. Random Forest and Decision Trees improve model interpretability, while Isolation Forest boosts anomaly detection abilities. The Convolutional Neural Network integration aids in extracting spatial features from input data, while the Hybrid GRU-CNN enhances the model’s capacity to capture temporal dependencies in cybersecurity incidents. The hybrid approach demonstrated superior performance, achieving an impressive accuracy of 98.86% in incident detection and response, as shown by experimental results. The proposed model shows significant progress in managing spatial and temporal aspects of cybersecurity data, surpassing individual models and conventional methods. Combining machine learning methods creates a strong base for developing flexible and durable AI-powered incident response systems in the changing field of Cybersecurity. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Cybersecurity

来源：评论

学校读者我要写书评

暂无评论

On Importance of Layer Pruning for Smaller BERT Models and Low Resource Languages

arXiv

引用

arXiv 2025年

作者： Shirke, Mayur Shembade, Amey Wagh, Madhushri Thorat, Pavan Joshi, Raviraj Pune Institute of Computer Technology Pune India L3Cube Labs Pune Maharashtra Pune India Indian Institute of Technology Madras Chennai India

This study explores the effectiveness of layer pruning for developing more efficient BERT models tailored to specific downstream tasks in low-resource languages. Our primary objective is to evaluate whether pruned BERT models can maintain high performance while reducing model size and complexity. We experiment with several BERT variants, including MahaBERT-v2 and Google-Muril, applying different pruning strategies and comparing their performance to smaller, scratch-trained models like MahaBERT-Small and MahaBERT-Smaller. We fine-tune these models on Marathi datasets, specifically Short Headlines Classification (SHC), Long Paragraph Classification (LPC) and Long Document Classification (LDC), to assess their classification accuracy. Our findings demonstrate that pruned models, despite having fewer layers, achieve comparable performance to their fully-layered counterparts while consistently outperforming scratch-trained models of similar size. Notably, pruning layers from the middle of the model proves to be the most effective strategy, offering performance competitive with pruning from the top and bottom. However, there is no clear winner, as different pruning strategies perform better in different model and dataset combinations. Additionally, monolingual BERT models outperform multilingual ones in these experiments. This approach, which reduces computational demands, provides a faster and more efficient alternative to training smaller models from scratch, making advanced NLP models more accessible for low-resource languages without compromising classification accuracy. Copyright © 2025, The Authors. All rights reserved.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Analysing the Relationship Between CSR and Financial Performance Using AI and Machine Learning Techniques 1

Analysing the Relationship Between CSR and Financial Perform...

引用

1st International Conference on Advances in computer Science, Electrical, Electronics, and Communication Technologies, CE2CT 2025

作者： Moholkar, Nilambari Duggal, Meenakshi Ingale, Devyani Awate, Sadhana Ganesh Raje, Harshal Dande, Mangesh Purushottam Dr.D.Y.Patil Institute of Technology Pimpri Pune India JSPM's Rajarshi Shahu college of Engineering Pune India RMDSSOMS Pune India Sinhgad Institute of Business Administration and Computer Application Lonavala India Global Business School and Research Centre Dr. D.Y Patil Vidyapeeth Maharashtra Pimpri Pune India Indira School of Business Studies PGDM Pune India

ISBN: (纸本)9798331518578

Corporate Social Responsibility (CSR) has arisen as a significant subject in contemporary business, with many organizations carrying out CSR techniques to fulfil partner assumptions and work on their standing. But since the monetary impacts of CSR drives are as yet indistinct, scholastics are investigating the way in which they connect with monetary execution. With regards to overseeing unpredictable, non-direct associations, conventional insightful procedures are restricted. New advancements in computerized reasoning (simulated intelligence) as well as AI (ML) give solid instruments to more exact examination of CSR and monetary execution. With an accentuation on anticipated exactness, cognizance, and reasonable consequences, this study investigates the manners by which man-made intelligence and ML approaches may be utilized to look at the association between monetary execution and CSR drives. In every industry, the Keras-trained MLP consistently yielded favourable results. The target of 900 epochs, or MLP (900), was used to train the MLP. It illustrates how training the same model over prolonged periods of time greatly improves performance. The LSTM model does not seem to generate better results, despite showcasing exceptional accuracy in natural language processing and long log security-related problems. As per the discoveries, artificial intelligence driven examination can give definite bits of knowledge into how corporate social responsibility (CSR) influences firm productivity, helping organizations in going with informed choices on their CSR strategies. © 2025 IEEE.

关键词： Artificial Intelligence CSR Financial Performance Machine Learning

来源：评论

学校读者我要写书评

暂无评论

Classification Models of Music Recommendation System for Mental Health Improvement: A Comparative Study

Classification Models of Music Recommendation System for Men...

引用

AIML-Applications for Engineering & technology (ICAET), International Conference on

作者： Aarti Amod Agarkar Kalyani Ghuge Shriram Dindore Suvidha Dhakne Atharva Dhakate Arya Mane Computer Engineering Department Vishwakarma Institute of Technology Pune India CSE - AIML Department Vishwakarma Institute of Technology Pune India

ISBN: (数字)9798350355611

ISBN: (纸本)9798350355628

Artificial Intelligence (AI) became part of our daily life. Due to the increased capability of computation power, AI significantly impacts human life and changing the social systems as well. On one side AI making our life easier but at the same time we are facing challenges like mental health issues. Lot of people are facing anxiety, insecurity and depression. Music is one of the remedy for improving mental health. In this paper, we focus on the music recommendation system for mental health improvement. The input imbalanced data is balanced to reduce its imbalance ratio. The output dataset is used to train the classification models including Random Forest, Support Vector Machine (SVM), Decision Trees, logistic regression, XGBOOST and ADABOOST. The comparative results are presented based on the experimentation. Results shows that XGBOOST outperforms other classification models with an accuracy of 98.1%.

关键词： Support vector machines Logistic regression Accuracy Anxiety disorders Mental health Depression Data models Decision trees Recommender systems Random forests

来源：评论

学校读者我要写书评

暂无评论

LexiEase AI : An AI Powered Dyslexic Support System

LexiEase AI : An AI Powered Dyslexic Support System

引用

Trends in Material Science and Inventive Materials (ICTMIM), International Conference on

作者： Sheetal Phatangare Sneha Jain Siddhartha Chakrabarty Shrey Chougule Somrath Bisen Department of Computer Engineering Vishwakarma Institute of Technology Pune India

ISBN: (数字)9798331501488

ISBN: (纸本)9798331501495

The development of an intelligent, dyslexia-specific support system is the need of the hour to ensure inclusion and personalization. The work presents an artificial intelligence-driven system, which involves state-of-the-art machine learning and NLP techniques, for identifying and assisting dyslexics. Our system encompasses a broad range of dyslexia screening tests, such as phonological awareness tests, memory tests, and oral reading, for which dyslexics may be tested using state-of-the-art models like DistilBERT and DeepSpeech. A Random Forest classifier generates certain dyslexia score to classify the severity levels and give customized learning pathways. The system will also integrate document simplification, mind map generation, and writing assistance features in order to make it holistic.

关键词： Human computer interaction Materials science and technology Adaptation models Writing Dyslexia Market research Data models Usability Random forests Testing

来源：评论

学校读者我要写书评

暂无评论

Novel Distributed Approach for Next Best Offer using XGboost

Novel Distributed Approach for Next Best Offer using XGboost

引用

Emerging Smart Computing and Informatics (ESCI), Conference on

作者： Shatakshi Chaudhari Prerana Malladi Sai Ashwini D Bundele Department of Computer Engineering Pune Institute of Computer Technology Pune India Department of Electronics and Communications Engineering Sardar Vallabhbhai National Institute of Technology Surat India

ISBN: (数字)9798331515683

ISBN: (纸本)9798331515690

Product recommendations have become integral to business development and customer engagement strategies. While traditional techniques like market basket analysis are effective in retail, they are insufficient for the financial sector due to the diversity of banking products and the lack of direct dependency among them. This research paper proposes a novel distributed framework for developing an automated recommendation system leveraging XGBoost to predict banking product purchases based on categorized customer profiles. The framework integrates distributed computing technologies such as Hadoop for data storage and PySpark for scalable data processing. The system classifies customers into distinct categories based on their transaction history, demographic attributes, and banking behaviors to deliver personalized product suggestions. Model performance is evaluated using metrics such as the ROC curve, precision, recall, and F1-score. Empirical results demonstrate the system’s effectiveness in enhancing lead generation, targeting high-potential customers, and improving customer satisfaction in real-world financial markets.

关键词： Measurement Technological innovation Memory Banking Lead History Ensemble learning Informatics Recommender systems Business

来源：评论

学校读者我要写书评

暂无评论

AlertCam-An Effective Anomaly Detection System using computer Vision

AlertCam-An Effective Anomaly Detection System using Compute...

引用

Emerging Smart Computing and Informatics (ESCI), Conference on

作者： Santushti Betgeri Piyush Lodha Aneesh Oak Aryan Parikh Ayush Lodha Dept. Of Computer Engineering Vishwakarma Institute of Technology Pune

ISBN: (数字)9798331515683

ISBN: (纸本)9798331515690

With the rising incidents of harassment and violence in India, particularly against women, this paper presents a real-time violence detection system using YOLO image detection models (YOLOv7, YOLOv8, and YOLOv9). The models were trained on a carefully curated and manually annotated dataset using Roboflow, ensuring robust detection across diverse scenarios. Transfer learning and hyperparameter tuning were applied to improve accuracy. The system was evaluated using key performance metrics, including mean average precision (mAP), Intersection over Union (IoU), precision, and recall, achieving 80% precision and 50–60% recall. Extensive manual testing on real-world surveillance footage and additional video sources further validated its effectiveness. Results indicate that YOLOv8 outperforms other versions in balancing speed and accuracy. This system has strong potential for deployment in law enforcement, public surveillance, and safety-critical applications.

关键词： YOLO Accuracy Surveillance Scalability Transfer learning Real-time systems Safety Tuning Standards Testing

来源：评论

学校读者我要写书评

暂无评论

Image Tampering Detection

Image Tampering Detection

引用

Computational, Communication and Information technology (ICCCIT), International Conference on

作者： Sandeep Shinde Apurv Waghmare Anushka Varpe Durva Ajgaonkar Arya Alurkar Department Of Computer Engineering Vishwakarma Institute Of Technology Pune India

ISBN: (数字)9798331512965

ISBN: (纸本)9798331512972

This paper proposes an image forgery detection technique combining the use of Convolutional Neural Networks (CNNs) with Error Level Analysis (ELA) for efficiently distinguishing tampered images from those that are original. ELA identifies potential regions of manipulation based on the discrepancy of the compression artifacts, whereas the VGG16based CNN classifier distinguishes between original and tampered images. The proposed model was assessed using the CASIA2 dataset, thus illustrating its effectiveness in accurately classifying images. This research contributes to the field of digital forensics, providing an advanced method for ensuring the integrity and authenticity of digital content, which is crucial for security and verification in various applications.

关键词： Measurement Visualization Accuracy Scalability Digital forensics Solids Forgery Convolutional neural networks Security Research and development

来源：评论

学校读者我要写书评

暂无评论

Decision Support System for Critical Cyber Security Infrastructure: AI and ML Approach 12th

Decision Support System for Critical Cyber Security Infrastr...

引用

12th International Conference on Frontiers of Intelligent Computing: Theory and Applications, FICTA 2024

作者： Shambharkar, Shubhangi Kumar, Anil Dari, Sukhvinder Singh Joshi, Ira Bhoite, Sonali Prashant Shirkande, Aparna S. Indurkar, Ira Computer Technology Department Yeshwantrao Chavan College of Engineering Maharashtra Nagpur India Department of Artificial Intelligence and Data Science Poornima Institute of Engineering and Technology Rajasthan Jaipur India Pune India Electronics and Communication Engineering Arya College of Engineering Rajasthan Jaipur India Department of Information Technology Vishwakarma Institute of Information Technology Pune Maharashtra Pune India S. B. Patil College of Engineering Indapur Maharashtra Pune India Department of Neuroscience DMCOP Datta Meghe College of Physiotherapy Maharashtra Nagpur India

ISBN: (纸本)9789819601462

The overall increase in cyber-attacks which takes place has had a major impact on the working of an organization and its related forensics department. With increased amount of breaches, the technical aspects of a developer are at stake. This had therefore led to the evolution of innovation and automation in the security domain. However, the risks which are associated with the same remains stagnant;since the digital transformation have been on a constant improvization factor with respect to the strategies thus involved. In such a scenario, the usage and implementation and artificial intelligence (AI) becomes a mandatory task. Cyber-attacks and breaches which are based on AI concepts have had a major impact and have revolutionized the overall working of the networking sector. An increase in the cyber-attack surface has been observed vividly. With this increase in the resilience of the networking sector, the author of the proposed research paper tends to highlight the need for AI- and ML-based decisive strategies which must be amalgamated in cyber security. The study also aims to focus on the role of various cyber forensic departments that have been successfully adopted by multiple companies. An understanding and working of a cyber-criminal is also needed in such a scenario. Hence, the knowledge of cyber security along with cyber forensic appears to be fruitful. The research paper thereby presents a systematic review of AI amalgamated with cyber forensics and provides various problem statements and methodologies associated with the same. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： AI Cyber forensics Methodologies ML Resilience

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：