Generative AI (GAI) can enhance the cognitive, reasoning, and planning capabilities of intelligent modules in the Internet of Vehicles (IoV) by synthesizing augmented datasets, completing sensor data, and making seque...
详细信息
Research in mobile robotics is growing into applications for difficult-to-access environments, such as in rescue and transport missions. Furthermore, autonomous vehicles can perform data collection in complex approach...
详细信息
Physico-mechanical properties of rocks have a direct correlation with the drilling rate of percussive drill. The prediction of drilling rate is important for the deployment of drills during the planning stage. In trop...
详细信息
Metaverse as the next-generation Internet provides users with physical-virtual world interactions. To improve the quality of immersive experience, users access to Metaverse service providers (MSPs) and purchase bandwi...
详细信息
Deep neural networks (DNNs) are widely used in fields like computer vision and natural language processing. A key component of DNN training is the optimizer. SGD-Momentum is popular in many DNN methodologies, such as ...
详细信息
ISBN:
(数字)9798331527471
ISBN:
(纸本)9798331527488
Deep neural networks (DNNs) are widely used in fields like computer vision and natural language processing. A key component of DNN training is the optimizer. SGD-Momentum is popular in many DNN methodologies, such as ResNet and DenseNet, due to its simplicity and effectiveness. However, its slow convergence rate limits its use. To overcome this, we introduce inter-gradient collision into SGD-Momentum, inspired by the elastic collision model in physics. This new method, called ICSGD-Momentum, aims to improve convergence. We provide theoretical proof of convergence and establish a regret bound for ICSGD-Momentum. Experiments on benchmarks including function optimization, CIFAR-100, ImageNet, Penn Treebank, COCO, and YCB-Video show that ICSGD-Momentum accelerates training and enhances the generalization performance of DNNs compared to optimizers like SGD-Momentum, Adam, Radam, Adabound, and AdaBelief.
Optimizing various wireless user tasks poses a significant challenge for networking systems because of the expanding range of user requirements. Despite advancements in Deep Reinforcement Learning (DRL), the need for ...
详细信息
ISBN:
(数字)9798350361261
ISBN:
(纸本)9798350361278
Optimizing various wireless user tasks poses a significant challenge for networking systems because of the expanding range of user requirements. Despite advancements in Deep Reinforcement Learning (DRL), the need for customized optimization tasks for individual users complicates developing and applying numerous DRL models, leading to substantial computation resource and energy consumption and can lead to inconsistent outcomes. To address this issue, we propose a novel approach utilizing a Mixture of Experts (MoE) framework, augmented with Large Language Models (LLMs), to analyze user objectives and constraints effectively, select specialized DRL experts, and weigh each decision from the participating experts. Specifically, we develop a gate network to oversee the expert models, allowing a collective of experts to tackle a wide array of new tasks. Furthermore, we innovatively substitute the traditional gate network with an LLM, leveraging its advanced reasoning capabilities to manage expert model selection for joint decisions. Our proposed method reduces the need to train new DRL models for each unique optimization problem, decreasing energy consumption and AI model implementation costs. The LLMenabled MoE approach is validated through a general maze navigation task and a specific network service provider utility maximization task, demonstrating its effectiveness and practical applicability in optimizing complex networking systems.
Branch-and-bound is a typical way to solve combinatorial optimization problems. This paper proposes a graph pointer network model for learning the variable selection policy in the branch-and-bound. We extract the grap...
详细信息
To increase programming productivity, developers often copy and paste the source code with or without changing it. However, they may also introduce significant downsides in the long run, including complicating the sof...
To increase programming productivity, developers often copy and paste the source code with or without changing it. However, they may also introduce significant downsides in the long run, including complicating the software and raising maintenance costs. The activity of duplicating the code is known as code cloning. They are classified into four types - Type-1, Type-2, Type-3, and Type-4. In this paper, the author presents a machine-learning approach for detecting code clones of all kinds except for Type-2. Abstract Syntax Trees are used to extract features from the methods. A distance combination approach combines two feature vectors of a pair of methods and their class labels. Once the dataset is finalised, a machine- learning approach is utilised to classify the clone type. Moreover, boosting classifiers like XGBoost, CatBoost, LightGBM, Gradient Boosting and AdaBoost are evaluated for the highest classification accuracy. From the results obtained, LightGBM outperformed all the other classifiers with the highest F1 score of 0.81. This study would motivate future researchers to focus on identifying the Type-2 clones and extracting novel features in determining the clone types.
Teeth segmentation and recognition are critical in various dental applications and dental diagnosis. Automatic and accurate segmentation approaches have been made possible by integrating deep learning models. Although...
详细信息
The sample selection approach is very popular in learning with noisy labels. As deep networks "learn pattern first", prior methods built on sample selection share a similar training procedure: the small-loss...
详细信息
暂无评论