Crowd simulation algorithms based on the social force model are widely used to simulate the movement behavior of real crowds due to their realism and flexibility. However, when modeling the behavior of high-density cr...
详细信息
ISBN:
(数字)9798350386776
ISBN:
(纸本)9798350386783
Crowd simulation algorithms based on the social force model are widely used to simulate the movement behavior of real crowds due to their realism and flexibility. However, when modeling the behavior of high-density crowds, the interacting forces between individuals become extraordinarily complex, which leads to individual pedestrians being prone to jerks and local abnormal behavior. To address this problem, this paper proposes a crowd simulation algorithm that combines a social force model and a consistency algorithm to constrain the pedestrians as a whole through the consistency algorithm and eliminate possible local abnormal behaviors. Ultimately, the efficacy of our algorithm was corroborated through a series of self-organizing simulation experiments.
the growth and development of internet technology have made life a lot better and easier. It has increased the convenience of performing essential transactions withthe click of a button using smart devices. Convenien...
详细信息
this paper focuses on the combination of ML algorithms for predictive maintenance through Internet of things (IoT)-based smart safety helmets in order to increase not only the level of safety operations and but as wel...
详细信息
ISBN:
(数字)9798350361155
ISBN:
(纸本)9798350361162
this paper focuses on the combination of ML algorithms for predictive maintenance through Internet of things (IoT)-based smart safety helmets in order to increase not only the level of safety operations and but as well equipment life in industrial settings. We study the present helmet technologies and reveal the weaknesses by tomorrow foreseeing technologies which exist in current predictive maintenance strategies. Proposing a new technique, ML algorithms suite is applied as an analyzing tool to collected data from sensors within the helmets. the system architecture is real-time data processed by an IoT framework to predict and act on maintenance needs up front. the results of the experimental validation confirm the implementation of our approach, that shows the significant increase of the predictive accuracy and timeliness as compared to the traditional methods. the research has not only enabaled to create smart safety equipment but also a scalable approach that could be maximized across other Internet of things applications.
the proceedings contain 17 papers. the topics discussed include: detecting falling rocks by estimating excavation points using laser pointer;a real-time target tracking algorithm based on improved kernel correlation f...
ISBN:
(纸本)9781665424257
the proceedings contain 17 papers. the topics discussed include: detecting falling rocks by estimating excavation points using laser pointer;a real-time target tracking algorithm based on improved kernel correlation filter;constant false alarm OMP algorithm based on sparse ISAR imaging of stationary aircraft target;detection of anal sac disease, Atresia Ani and rectal prolapse for Canis lupus familiaris using image processing and convolutional neural network;investigation of gas/oil/water distribution based on electrical capacitance and microwave tomography;lung nodule classification of CT images based on the deep learning algorithms;the research on the impact of visual image perception in a user interface on typeface design;and joint robust transmit and receive beamforming based on probability constraint.
the proceedings contain 17 papers. the special focus in this conference is on Financial Technology. the topics include: Improved Machine Learning algorithms for Fraud Detection in Fintech Companies;A Cognitive Analysi...
ISBN:
(纸本)9789819638109
the proceedings contain 17 papers. the special focus in this conference is on Financial Technology. the topics include: Improved Machine Learning algorithms for Fraud Detection in Fintech Companies;A Cognitive Analysis of CEO Speeches and their Effects on Stock Markets;does E-Money Mediate the Effect of fundamental Factors on the Stock Price Index?;Multi-Attribute Decision Making (MADM) Model Based on Bayesian Neural Network Classification with Comparative Scoring Quantification Method Can Solve Intangible Value Valuation Issues in Benefit-Cost Analysis (BCA);Exploring Vulnerabilities in Near Field Communication (NFC) Devices: A Comprehensive Investigation;a Novel Electronic Payment System Based on Zero-Knowledge Proof and Blockchain;A Comprehensive AI and Blockchain Framework for Detecting and Preventing Money Laundering in Bangladesh Financial Systems;detecting Persuasion in Financial Short Texts: A Computational Approach;the Impact of Colored Noise on the CIR Model;A Secure NFC Cardless Cash Withdrawal System;FinTech Digital Transformation: Generative AI, Humanoid Robots, Metaverse, Human-AI Collaboration, and Industry 5.0;Artificial Intelligence (AI) and Virtual Reality Convergence in Financial Services: the Power of Digital Twin Robo-Advisers;CITRONN: A Convolutional Neural Network for Crypto Image-Based Trading;can the Government’s Distribution of Consumption Vouchers Stimulate Adoption of Digital Payment Channels? Insights from a Social Learning Perspective;financial Time Series Simulation with Transformer-Based Generative Models Under Continuous Conditions.
In order to promote smooth cross-cultural communication, this research focuses on improving Natural Language Processing (NLP) features within multilingual Chabots. the need for efficient communication across various l...
In order to promote smooth cross-cultural communication, this research focuses on improving Natural Language Processing (NLP) features within multilingual Chabots. the need for efficient communication across various linguistic and cultural barriers has grown as the world becomes more interconnected. In this context, multilingual Chabot's are essential tools that facilitate easy communication between users of various cultural backgrounds. In order to better understand and respond to a wide range of linguistic details, idioms, and cultural allusions, the research explores ways to improve NLP algorithms. this research investigates how machine learning techniques can be integrated to improve the Chabot's ability to adapt to various linguistic patterns, enabling it to acquire new language skills over time. By implementing an extensive and flexible approach, the proposed approach seeks to transform Natural Language Processing in Multilingual Chatbots for Cross-Cultural Communication. In order to identify language-specific patterns and specifications, the methodology starts withthe collection of a diverse dataset that is representative of linguistic variations and cultural contexts. It then uses advanced linguistic analysis techniques like phonetic analysis. the next step is cultural context modelling, which entails building a database enhanced with cultural allusions and language variations unique to a given context that are obtained from various sources. Transfer learning, reinforcement learning for dialogue, and modern algorithms such as Multilingual Unsupervised and Supervised Embeddings (MUSE) for cross-lingual embeddings are all included in machine learning integration. In order to obtain information about language precision and cultural sensitivity, the methodology incorporates interactive surveys and sentiment analysis into a strong user feedback mechanism. the proposed model achieved 98% performance.
the proceedings contain 12 papers. the topics discussed include: performance evaluation of some adaptive task allocation algorithms for fog networks;exploring task placement for edge-to-cloud applications using emulat...
ISBN:
(纸本)9781665402910
the proceedings contain 12 papers. the topics discussed include: performance evaluation of some adaptive task allocation algorithms for fog networks;exploring task placement for edge-to-cloud applications using emulation;AVEC: accelerator virtualization in cloud-edge computing for deep learning libraries;priority-enabled load balancing for dispersed computing;multilayer resource-aware partitioning for fog application placement;CHANGE: delay-aware service function chain orchestration at the edge;LEAF: simulating large energy-aware fog computing environments;reducing the mission time of drone applications through location-aware edge computing;TOD: transprecise object detection to maximize real-time accuracy on the edge;mapping IoT applications on the edge to cloud continuum with a filter stream mode;and PA-offload: performability-aware adaptive fog offloading for drone image processing.
the purpose of this paper is to realize the path planning of the search and rescue process of the UAV at sea. the first part of this paper wonder to find local optimal path planning from the UAV launch base station to...
the purpose of this paper is to realize the path planning of the search and rescue process of the UAV at sea. the first part of this paper wonder to find local optimal path planning from the UAV launch base station to the target area. Meanwhile the next part is to find full-coverage optimal path planning of the UAV in the target area, and to ensure the minimum energy consumption in the process of achieving the goal. For local path planning, this paper adopts the A* algorithm as the main algorithm and integrates the DWA algorithm to improve it, for the purpose to reduce the number of UAV turns and turn angles while achieving dynamic obstacle avoidance during the flight. For full-range path planning, this paper adopts the RCPP algorithm to achieve full-area coverage of the UAV on various irregular sea areas, and the energy efficiency of the UAV flight can be improved by determining the best starting and ending points. through the implementation of the two-part algorithm, it can meet the speed and accuracy requirements of maritime UAV search and rescue, ensure the shortest search and rescue time, and achieve the most efficient energy utilization.
Cardiovascular disease remains one of the leading causes of death worldwide, necessitating the development of efficient diagnostic tools. this study presents a machine learning framework for predicting cardiovascular ...
详细信息
ISBN:
(数字)9798331530983
ISBN:
(纸本)9798331530990
Cardiovascular disease remains one of the leading causes of death worldwide, necessitating the development of efficient diagnostic tools. this study presents a machine learning framework for predicting cardiovascular disease based on clinical and demographic data. We employ a comprehensive data preprocessing pipeline that includes handling missing values, normalizing data, and balancing the dataset to ensure robust model performance. Feature extraction and selection techniques are applied to identify the most relevant predictors of cardiovascular risk, optimizing model performance and reducing computational complexity. Eight machine learning algorithms were employed to predict cardiovascular disease outcomes, including logistic regression, decision trees, random forest, support vector machines, k-nearest neighbors, naïve Bayes, XGBoost, and AdaBoost model. Our approach achieved an accuracy of surpassing 98%, demonstrating the potential of ML techniques in aiding early diagnosis and improving patient outcomes. this comparative analysis highlights the strengths and limitations of each algorithm, providing insights into the most suitable models for clinical use.
the proceedings contain 14 papers. the topics discussed include: a comparison of nearest distance optimization and ant colony optimization for order picking in a multi-aisle warehouse;automatic test case generation me...
ISBN:
(纸本)9781450389679
the proceedings contain 14 papers. the topics discussed include: a comparison of nearest distance optimization and ant colony optimization for order picking in a multi-aisle warehouse;automatic test case generation method based on improved whale optimization algorithm;negative learning in ant colony optimization: application to the multi-dimensional knapsack problem;biased random-key genetic algorithms using path-relinking as a progressive crossover strategy;hand crafted and learned spatiotemporal filters to inform and track visual saliency;feature selection and feature extraction: highlights;and cross-domain recommendation based on heterogeneous information network with adversarial learning.
暂无评论