检索结果-内蒙古大学图书馆

39th IEEE International Conference on Data engineering, ICDE 2023

作者： Jin, Jipeng Lu, Guangben Li, Sijia Gao, Xiaofeng Tan, Ao Wang, Lifeng Shanghai Jiao Tong University MoE Key Lab of Artificial Intelligence Dept. of Computer Science and Engineering China Corporate Development Group Tencent China

ISBN: (纸本)9798350322279

Estimating conversion rate (CVR) accurately has been one of the most central problems in online advertising. Existing methods in production focus on learning effective interactions among features to boost the model performance. Despite great success, these methods treat all the features equally without distinction. However, different features suffer differently from cold-start issues. Tail elements in those high-cardinality features, which we denote as fine-grained features, tend to have inadequate samples and thus fail to obtain semantically meaningful embeddings. Interacting with those features leads astray and impairs the accuracy of new ads in a cold-start scenario. In this paper, we propose Automatic Fusion Network (AutoFuse) to better tackle the challenge. AutoFuse explicitly separates features into groups based on their granularity and learns multiple levels of representation conditioned on different combinations of feature groups. Concretely, AutoFuse learns an ad-level representation to depict the unique individual character and a group-level representation to portray the collective information by discarding the fine-grained features. The final robust and general ad representation is obtained by integrating these two level representations adaptively. Such a combination encompasses a wider amount of information, and thereby mitigates the cold-start issue. Extensive experiments on two industrial-scale datasets and three public datasets show that AutoFuse significantly and consistently outperforms a spectrum of competitive methods including our currently deployed model. Meanwhile, the remarkable improvement on new ads validates the effectiveness of our method in cold-start scenarios. We design AutoFuse as a generic approach and thus it can be seamlessly transferred into other domains. Our method has been deployed online to serve billions of users and ads and has achieved significant GMV gain of 2.84%. © 2023 IEEE.

关键词： Marketing

来源：评论

学校读者我要写书评

暂无评论

A Multi-Scale Learning Assisted Neural Network for Improved Object Detection in Complex Environments 9

A Multi-Scale Learning Assisted Neural Network for Improved ...

引用

9th International Conference on Communication and Electronics Systems, ICCES 2024

作者： Karthikayani, K. Vadlamani, Subramanyam M. Prasad, M.V.Siva Thatipudi, Jayendra Gopal Kalaivaani, P.C.D. Upadhyay, Shrikant Srm Institute of Science and Technology Chennai India Anurag engineering college Department of Cse Kodad India University of North Texas Dept. of Computer Science United States Kongu Engineering College Dept of Cse Perundurai638060 India Mlr Institute of Technology Department of Ece Hyderabad India

ISBN: (纸本)9798350377972

Object detection in complex environments presents significant challenges due to the variability in object scales, occlusions, and cluttered backgrounds. This paper proposes a Multi-Scale Learning Assisted Neural Network designed to improve object detection accuracy in such scenarios. The proposed model integrates multi-scale learning, hybrid attention mechanisms, and an advanced feature fusion strategy to enhance its ability to detect objects across various scales and complexities. The model was evaluated against nine state-of-the-art object detection models, including YOLOv4, Faster R-CNN, SSD, EfficientNet, RetinaNet, CenterNet, Cascade R-CNN, FCOS, and Mask R-CNN. Experimental results demonstrate that the proposed model outperforms existing models, achieving an accuracy of94.25%, precision of93.50%, recall of92.70%, and an F1-score of93.10%. Additionally, the model exhibits a balanced trade-off between inference time and model size, making it suitable for real-world applications where performance and resource efficiency are critical. The high IoU (87.50%) and mAP (89.20%) scores further validate the model's robustness in accurately detecting and localizing objects in complex environments. This study highlights the potential of the proposed model to significantly improve object detection tasks, particularly in challenging and dynamic real-world scenarios. © 2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

In-memory Machine Learning using Adaptive Multivariate Decision Trees and Memristors

In-memory Machine Learning using Adaptive Multivariate Decis...

引用

IEEE International Symposium on Circuits and Systems (ISCAS)

作者： Akash Chavan Pranav Sinha Sunny Raj Dept. of Computer Science and Engineering Oakland University Rochester MI USA

ISBN: (数字)9798350330991

ISBN: (纸本)9798350331004

We introduce a framework to design in-memory decision tree machine-learning (ML) circuits using memristor crossbars. Decision trees (DTs) offer many advantages over neural networks, such as enhanced energy efficiency, interpretability, safety, privacy, and speed, along with reduced dependence on extensive training data. We propose an adaptive multivariate decision tree (AMDT) training algorithm, which constructs decision trees that incorporate both univariate and multivariate features, facilitating the creation of higher accuracy and energy-efficient crossbar designs compared to the state-of-the-art (SOTA). Our circuits are realized using pure memristor crossbars, requiring just one memristor per cell and no transistors while employing sneak-paths for flow-based in-memory computations. In comparison to the SOTA, our approach produces designs that are, on average, 4% more accurate and require 12.6% lower energy.

关键词： Training Accuracy Circuits Memristors Training data Machine learning Energy efficiency

来源：评论

学校读者我要写书评

暂无评论

Multivariate Analysis of Urban Air Pollution: Clustering and Patterns Across Major Asian Cities

Multivariate Analysis of Urban Air Pollution: Clustering and...

引用

Future Machine Learning and Data science (FMLDS), IEEE International Conference on

作者： Anika Rahman Mst. Taskia Khatun Dept. of Computer Science & Engineering Stamford University Bangladesh Dhaka Bangladesh

ISBN: (数字)9798350391213

ISBN: (纸本)9798350391220

This study analyzes air pollution in Asian cities using the Global Air Pollution Data, consisting of 6,196 entries from 31 countries. Our primary goal is to identify pollution patterns through multivariate analysis and evaluate the effectiveness of six clustering algorithms: K-Means, Hierarchical Clustering, DBSCAN, Gaussian Mixture Models (GMM), Agglomerative Clustering, and Spectral Clustering. Performance was assessed using Silhouette Score, Davies-Bouldin Index, Calinski-Harabasz Index, WCSS, Cohesion, and Separation. The novelty of this work lies in the comparative analysis of these clustering methods on air pollution data, providing new insights into pollution dynamics across Asian cities. The analysis identified four distinct clusters- ‘High Pollution’, ‘Moderate Pollution’, ‘Ozone-Dominated Pollution’, and ‘Low Pollution’- with K-Means proving to be the most effective. Significant disparities were found, particularly in South and East Asia, where countries like India, China, and Pakistan exhibited the highest pollution levels. Additionally, an examination of capital cities revealed specific pollution patterns and the primary pollutants-PM2.5, NO2, CO, and Ozone-aiding in identifying sources and affected regions. These findings underscore the need for targeted regional pollution control strategies.

关键词： Gases Environmentally friendly manufacturing techniques Asia Urban areas Pollution control Clustering algorithms Transportation Air pollution Indexes Public healthcare

来源：评论

学校读者我要写书评

暂无评论

COMPOSITIONAL GENERATIVE INVERSE DESIGN

arXiv

引用

arXiv 2024年

作者： Wu, Tailin Maruyama, Takashi Wei, Long Zhang, Tao Du, Yilun Iaccarino, Gianluca Leskovec, Jure Dept. of Engineering Westlake University China NEC Laboratories Europe Germany Dept. of Computer Science MIT United States Dept. of Mechanical Engineering Stanford University United States Dept. of Computer Science Stanford University United States

Inverse design, where we seek to design input variables in order to optimize an underlying objective function, is an important problem that arises across fields such as mechanical engineering to aerospace engineering. Inverse design is typically formulated as an optimization problem, with recent works leveraging optimization across learned dynamics models. However, as models are optimized they tend to fall into adversarial modes, preventing effective sampling. We illustrate that by instead optimizing over the learned energy function captured by the diffusion model, we can avoid such adversarial examples and significantly improve design performance. We further illustrate how such a design system is compositional, enabling us to combine multiple different diffusion models representing subcomponents of our desired system to design systems with every specified component. In an N-body interaction task and a challenging 2D multi-airfoil design task, we demonstrate that by composing the learned diffusion model at test time, our method allows us to design initial states and boundary shapes that are more complex than those in the training data. Our method generalizes to more objects for N-body dataset and discovers formation flying to minimize drag in the multi-airfoil design task. Project website and code can be found at https://***/AI4science-WestlakeU/cindm. © 2024, CC BY.

关键词： Airfoils

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of Synthetic Aperture Radar Images 8

Performance Evaluation of Synthetic Aperture Radar Images

引用

8th International Conference on Signal Processing and Communication, ICSC 2022

作者： Upadhyay, Kritika Kumar, Sandeep Bharti, Manisha Agarwal, Prabhakar National Institute of Technology Delhi Dept. of Electronics and Communication Engineering Delhi India National Institute of Technology Delhi Dept. of Computer Science and Engineering Delhi India

ISBN: (纸本)9781665454308

Synthetic Aperture Radar (SAR) imaginary is used extensively for Military applications in the modern era of technology. In this work, we have evaluated the performance of SAR images for Radar Systems. The SAR length, Resolution, and Footprint of SAR networks have been compared by varying the parameters of Radar signals like Bandwidth, Operating Frequency, and Satellite Altitude. From the results, it was observed that Antenna Footprints vary inversely with the Operating Frequency and Grazing Angle of Radar © 2022 IEEE.

关键词： Synthetic aperture radar

来源：评论

学校读者我要写书评

暂无评论

Efficient mmWave Beam Selection using ViTs and GVEC: GPS-based Virtual Environment Capture 14

Efficient mmWave Beam Selection using ViTs and GVEC: GPS-bas...

引用

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

作者： Gupta, Srajan Saxena, Navrati Roy, Abhishek Ravi, Akshay San Jose State University Dept. of Computer Science San Jose United States Mediatek Inc. Engineering San Jose United States

ISBN: (纸本)9798350335095

Millimeter waves (mmWaves) providing higher bandwidth is used by 5G network technology to achieve higher network capacity and faster data transfer. However, the process of beam sweeping across multiple antenna arrays can be time-consuming and inefficient. Machine learning (ML) models can address this issue by predicting the optimal beam pair based on data from on-vehicle light detection and ranging (LIDAR) and global positioning system (GPS) sensors. This paper proposes a new vision transformer (ViT) ML model for beam selection using GPS and LIDAR data. GPS-based virtual environment capture (GVEC) has been introduced to overcome to overcome noise in the LIDAR data caused by adverse weather conditions. The proposed solution demonstrates improved performance compared to previous approaches when tested on noisy LIDAR data, achieving a 92% accuracy in searching among the top 10 beams. © 2023 IEEE.

关键词： Millimeter waves

来源：评论

学校读者我要写书评

暂无评论

An algorithm frequently mines a set of items for big data to enhance efficiency

An algorithm frequently mines a set of items for big data to...

引用

Automation and Computation (AUTOCOM), International Conference on

作者： Sanjay Kumar Sonkar Balkar Singh Dept. of Computer Science and Engineering SunRise University Alwar Rajasthan India

In the landscape of rapidly expanding data streams, deriving meaningful insights, particularly frequent itemsets, within massive datasets at a swift pace poses a significant challenge. Addressing this challenge, various methodologies for Frequent Itemset Mining have been proposed, yet they struggle with low support counts and efficiency concerns in handling large datasets. In response, our research introduces Jagged Itemset Counting (JIC) methodologies, aiming to effectively mine Frequent Itemsets from extensive data. The core objective revolves around devising a robust algorithm capable of identifying all Frequent Itemsets, irrespective of database size or the nature of the itemset. Central to this approach is the introduction of a straightforward label representation, GPLN (Geometric Progression Label Number), assigned to each frequent item. Utilizing CGPLN (Cumulative Geometric Progression Label Number), derived from the arithmetic sum of GPLNs within transaction subsets, forms the CGPLN-Label representation for each transaction subset (itemset). Comparative analysis reveals superior performance of JIC over Apriori and Eclat algorithms for small and medium-sized datasets, exhibiting efficiency even at minimal support thresholds. In the realm of Big Data, where FP-Growth and Eclat falter, the proposed technique shines with faster execution times, optimized main memory utilization, and efficient disc memory usage.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Analyzing Yelp Open-Source Data-Set in Azure Data-bricks

Analyzing Yelp Open-Source Data-Set in Azure Data-bricks

引用

International Conference on Inventive Computation Technologies (ICICT)

作者： Mugunth D Maheshwaran R Maheswari M S. L. Jany Shabu J. Refonaa Dept. of Computer Science and Engineering Sathyabama Institute of Science and Technology Chennai Tamilnadu India

ISBN: (数字)9798350359299

ISBN: (纸本)9798350359305

A data-driven application, Yelp has always served and will always serve as one. It is one of the first companies that allows local businesses to receive reviews from their customers. Communities have always developed it in cooperation with each other. According to Yelp, the data-set contains information regarding businesses, reviews, users, and check-ins, and has continuously been updated since 2015. A project titled "Analyzing Yelp Open-Source Data-set in Azure Data Bricks" will analyze this data-set which has been made available as an open source for sentiment analysis and descriptive analysis. This study analyzes local business performance, business distribution, review ratings, and other factors, as well as check-in rates in American business locations over time. This analysis shows that the Yelp is losing reviews, tips, elite users, and check-ins.

关键词： Sentiment analysis Reviews Decision making Data visualization Machine learning Companies Data science

来源：评论

学校读者我要写书评

暂无评论

Smart Healthy Schools: An IoT-enabled concept for multi-room dynamic air quality control

引用

Internet of Things and Cyber-Physical Systems 2024年第1期4卷 24-31页

作者： Zivelonghi, Alessandro Giuseppi, Alessandro Dep. of Applied Mathematical Science IIS "Calabrese-Levi" VR San Pietro in Cariano Italy PowerConsulting Studio San Pietro in Cariano Verona Italy Dept. of Computer Control and Management Engineering University of Rome "La Sapienza" Via Ariosto 25 Rome00185 Italy

Smart Healthy Schools (SHS) are a new paradigm in building engineering and infection risk control in school buildings where the disciplines of Indoor Air Quality (IAQ), IoT (Internet of Things) and Artificial Intelligence (AI) merge together. In the post-pandemic era, equipping schools with a network of smart IoT sensors has become critical to aspire for the optimal control of the IAQ and lowering the airborne infection risk of several pathogens, indirectly related to cumulated human emitted CO2 levels over time. Thermal energy waste in winter due to improved air renewal remains of major concern but can be well monitored within a SHS monitoring architecture thanks to the flexibility of the LoRaWAN protocol able to process also a large amount of energy and climatic data at room and building scale. In this work, we report the design of the AulaSicura platform, an IoT control system co-designed by the main author and Gizero Energie to implement the SHS paradigm via clearly visible (and audible) alarm signalling in existing and new school buildings. The cloud-based LoRa system is capable of continuous and simultaneous monitoring of a variety of sensors and IAQ parameters including indoor/oudoor temperatures, rel. humidities and human-emitted excess CO2. The multi-room monitoring concept of indoor-CO2 levels allows centralized control of natural ventilation levels in individual classrooms and can handle (quasi)-real-time data, relevant for data post-processing and future developments in (quasi)-real-rime assessment of IAQ and infection risk levels at single room scale. The sensor network is also extensible to up to one thousand of classrooms per LoRa-node allowing centralized control of entire school districts at an urban scale. Moreover, through Modbus-LoRa I/O converters, AulaSicura can also control the same amount of mechanical ventilation units per node either in pure or hybrid mechanical ventilation modes. © 2023 The Authors

关键词： Internet of things

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：