The proceedings contain 135 papers. The topics discuss include: digital twins in smart farming;analyzing cybersecurity patterns in the pacific region: trends and challenges for 2023;transfer learning in Monte Carlo tr...
ISBN:
(纸本)9798350341072
The proceedings contain 135 papers. The topics discuss include: digital twins in smart farming;analyzing cybersecurity patterns in the pacific region: trends and challenges for 2023;transfer learning in Monte Carlo tree search;two-stage fine-tuning for low-resource English-based Creole with pre-trained LLMs;combining first-order and texture features with convolutional neural network for classification of residential area density in aerial images;shallow learning for predictive blood test anomaly detection: case study for rheumatic diseases;deep learning based intrusion detection for internet of things and edge devices;evaluation of data durability in erasure coding using peak shift method of drive failure risk;and review of secure healthcare and patient monitoring system using block chain for accessing medical records.
The proceedings contain 133 papers. The topics discussed include: autonomous UAV landing on mobile platforms;a comparative study of hyper-parameter optimization tools;empirical stability limits for a size-based schedu...
ISBN:
(纸本)9781665495523
The proceedings contain 133 papers. The topics discussed include: autonomous UAV landing on mobile platforms;a comparative study of hyper-parameter optimization tools;empirical stability limits for a size-based scheduler applied to network utility maximization;enhancing object clarity in single channel night vision images using deep reinforcement learning;a deep reinforcement learning scheduler with back-filling for high performance computing;predicting effectiveness of marketing through analyzing emotional context in advertisement using deep learning;churn prediction in telecom industry using machine learning ensembles with class balancing;deep learning method for haze prediction in Singapore;evaluating deep sequential knowledge tracing models for predicting student performance;and detection of android malware using tree-based ensemble stacking model.
The proceedings contain 78 papers. The topics discussed include: bimodal ERP study with auditory-visual stimuli;development of regenerative and low pressure drop adsorbent structure for biogas upgrading;advanced data ...
ISBN:
(纸本)9781728163031
The proceedings contain 78 papers. The topics discussed include: bimodal ERP study with auditory-visual stimuli;development of regenerative and low pressure drop adsorbent structure for biogas upgrading;advanced data balancing method with SVM decision boundary and bagging;long-term impact of non-egalitarian gender societal structures: a gendered game of life study;increasing hydrogen energy efficiency by heat integration between fuel cell, hydride tank and electrolyzer;effect of clay content on performance of a crushed rock mix;multi-agent simulation of relationality assets to enable community vitalization;using graph convolution network for predicting performance of automatically generated convolution neural networks;and practices of software testing techniques and tools in Bangladesh software industry.
The proceedings contain 123 papers. The topics discussed include: sizing and mission analysis of multirole short-haul all-electric seaplane for sustainable aviation;impact of business technologies on the success of e-...
ISBN:
(纸本)9781665419741
The proceedings contain 123 papers. The topics discussed include: sizing and mission analysis of multirole short-haul all-electric seaplane for sustainable aviation;impact of business technologies on the success of e-commerce strategies: SMEs perspective;estimating the size of crowds through deep learning;digital gender divide in online education during Covid-19 lockdown in India;towards an analytical probe for twitter information flow micro-structure;performance evaluation of permissioned blockchain platforms;application of hyperspectral imaging technique to determine the quality of photo and thermal exposed and contaminated pharmaceutical formulations: a cost effective way of quality testing;lepidoptera classification through deep learning;barriers for an integrated lean and ISO 14001 implementation for sustaining environmental performance in the manufacturing industry;and examination of the optimum regional point system by gift & circulation model.
The proceedings contain 118 papers. The topics discussed include: analysis support by estimating speaker's state in sensor-based active listening;assessing the suitability of three-point and four-point bending sim...
ISBN:
(纸本)9781665453059
The proceedings contain 118 papers. The topics discussed include: analysis support by estimating speaker's state in sensor-based active listening;assessing the suitability of three-point and four-point bending simulation to investigate the fracture mechanisms in directionally-reinforced fiber composites;solar irradiance prediction using deep learning-based approaches;automated assessment of excessive inbreeding of dogs based on deep learning;estimating creativity drawing features from hand drawing logs;understanding the learning of disabled students: an exploration of machine learning approaches;memory access characteristics of neural network workloads and their implications;providing automated feedback on the gestalt principle of symmetry by means of machine learning;comparative analysis of classification techniques on Olympic games datasets;a combination of technical indicators and deep learning to predict price trends for short-term cryptocurrency investment;orchestrating the resilience of cloud microservices using task-based reliability and dynamic costing;and a retrieval method of software requirements from Japanese requirements document with dependency analysis and keywords.
One of the key challenges in federated learning is addressing the non-independent and identically distributed (Non-IID) data among parties, which can lead to divergence of local model parameters and decreased converge...
详细信息
ISBN:
(纸本)9789819770069;9789819770076
One of the key challenges in federated learning is addressing the non-independent and identically distributed (Non-IID) data among parties, which can lead to divergence of local model parameters and decreased convergence accuracy of the global model, along with serious dimension collapse issues. In this paper, we introduce VICON (Variance-Invariance-Covariance model Contrastive Learning), a method to prevent dimension collapse issues. In local training, particularly, a regularization technique is employed to foster orthogonal feature representations across dimensions and to sustain the variance of individual embedded dimensions above a predefined level. Complementary to this, contrastive learning methodologies are utilized to cluster-like instances while dissociating divergent ones, further enhancing discriminative capabilities. It helps control the model's parameter norm and adapt it to high-dimensional data, reducing information loss and aligning local models with the global optimization objective in federated learning to minimize bias and collapse. Extensive experiments show that VICON performs better than other algorithms such as federated averaging (FedAvg), federated proximal optimization (FedProx), and model-agnostic federated learning (Moon). Compared to the Moon algorithm, it not only improves accuracy by 2.2% to 3.7%, but also enables efficient communication, achieves higher accuracy, and remains robust when dealing with imbalanced data and uncertain local updates.
In order to reduce the multiplicative depth required by high-order polynomial evaluation for homomorphic encrypted data, we propose two novel and low multiplicative depth polynomial evaluation algorithms: an x4-step n...
详细信息
This paper proposes a course intelligent recommendation system based on machine learning algorithm to achieve more accurate personalized recommendation. The system uses deep learning model to design an optimized recom...
详细信息
In computer Vision (CV), the deployment of Convolutional Neural Networks (CNNs) is often hindered by their substantial computational requirements and large labeled datasets. Self-supervised learning (SSL) serves as an...
详细信息
This paper explored the application of mmWave radar sensors in robotic navigation. With the growing need for indoor navigation and object avoidance employing mmWave technology, a mobile robot equipped with a mmWave ra...
详细信息
暂无评论