检索结果-内蒙古大学图书馆

13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)

作者： Wang, Bing Ma, Wenming Yantai Univ Sch Comp & Control Engn Yantai 264005 Peoples R China

ISBN: (纸本)9780738105451

With the rapid development of science and technology, huge amounts of information fill people's lives, and the accompanying information overload phenomenon has become an urgent problem to be solved. Because the recommendation system can quickly find the products users want in the massive item information, to a certain extent It has attracted much attention to solve the problem of information overload. Matrix factorization is a commonly used technique in recommendation systems. It can effectively improve the recommendation effect when the scoring matrix is sparse. However, due to its own reasons, matrix factorization has many problems such as sparseness, cold start, and low interpretability. In the field of deep learning, because the normalization technology BatchNorm can optimize the training process, accelerate the training speed and make the training results more stable, it has been studied by a large number of scholars. In this paper, Matrix Factorization Based on BatchNorm and Preference Bias is proposed. BatchNorm is combined with matrix factorization, user and item preferences are added, and adam algorithm is used for optimization. Experiments show that the algorithm in this paper has a good recommendation effect on sparse matrix.

关键词： matrix factorization batchnorm bias adam algorithm optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

CNN-Based Broad Learning System 4

CNN-Based Broad Learning System

引用

4th IEEE International Conference on Signal and Image Processing (ICSIP)

作者： Li, Ting Fang, Bin Qian, Jiye Wu, Xuegang Chongqing Univ Coll Comp Sci Chongqing Peoples R China State Grid Chongqing Elect Power Co Elect Power Res Inst Chongqing Peoples R China Yangtze Normal Univ Coll Comp Sci Chongqing Peoples R China

ISBN: (纸本)9781728136608

The Broad Learning System (BLS) has recently been proposed as an effective and useful method for pattern recognition. It overcomes the shortcomings of deep neural networks that are both time consuming and hardware dependent. BLS is a lightweight network structure. Although its advantages in speed and flexibility arc obvious, there is a still a gap between its accuracy in image recognition and that of state-of-art approaches. We propose a BLS-based model that combines CNN and adam algorithm. At first, we build a BLS-based classifier, and then we build a CNN-based feature extraction. After that, we take features extracted as inputs to the classifier. Feature extraction through convolution and pooling operations can effectively preserve key features of the image. We use adam algorithm to update the weights of the feature extraction. By retaining key features and removing factors that affect image recognition, it helps to improve the accuracy of the classification results. Therefore, through the above operations, the speed and accuracy of our proposed lightweight network structure haves been improved and reflected in the dataset of the MNIST and Cat or Dog for Kaggle competitions. We can choose different convolution layers, pooling layers and optimization algorithms according to the different sized data sets.

关键词： broad learning system convolutional neural network feature extraction adam algorithm

来源：评论

学校读者我要写书评

暂无评论

Development of Design Patterns with Adaptive User Interface for Cloud Native Microservice Architecture Using Deep Learning With IoT 5

Development of Design Patterns with Adaptive User Interface ...

引用

5th IEEE International Conference on Computing, Power, and Communication Technologies, IC2PCT 2024

作者： Cherukuri, Bangar Raju Germantown MD United States

ISBN: (纸本)9798350383522

The development of design patterns with adaptive user interfaces using deep learning provides an innovative solution to address the demands of modern software systems. The design patterns play an essential role in obtaining scalability and reliability in the systems. The various challenges in the architectural designs and automatic decision-making techniques are attained through the aid of adam optimization techniques integrated with the Internet of Things. Through the process of integration with an adaptive user interface, the changing user needs are monitored by the system. The deep learning techniques help to automatically adapt and optimize itself without any external parameters. This helps to analyze the user interactions and to promote system efficiency with an enhanced performance ratio. The generation of constant streams of data is obtained through the Internet of Things. The real-time decisions are adopted by analyzing the user behaviour. The adam optimization techniques are well-known for the training of neural networks. This helps to maintain and adopt fine-tuning of the system and maintaining optimum performance. Thus the integration of deep learning with cloud-native microstructure architecture provides self-optimizing networks for each user in the dynamic nature of the IoT data in real-time environments. © 2024 IEEE.

关键词： adam algorithm Adaptive user interface Decision-making Deep learning Design patterns Internet of Things Software systems

来源：评论

学校读者我要写书评

暂无评论

Advancing Supervised Learning with the Wave Loss Function: A Robust and Smooth Approach

引用

PATTERN RECOGNITION 2024年 155卷

作者： Akhtar, Mushir Tanveer, M. Arshad, Mohd. Indian Inst Technol Indore Dept Math Indore 453552 India

Loss function plays a vital role in supervised learning frameworks. The selection of the appropriate loss function holds the potential to have a substantial impact on the proficiency attained by the acquired model. The training of supervised learning algorithms inherently adheres to predetermined loss functions during the optimization process. In this paper, we present a novel contribution to the realm of supervised machine learning: an asymmetric loss function named wave loss. It exhibits robustness against outliers, insensitivity to noise, boundedness, and a crucial smoothness property. Theoretically, we establish that the proposed wave loss function manifests the essential characteristic of being classification-calibrated. Leveraging this breakthrough, we incorporate the proposed wave loss function into the least squares setting of support vector machines (SVM) and twin support vector machines (TSVM), resulting in two robust and smooth models termed as Wave-SVM and Wave-TSVM, respectively. To address the optimization problem inherent in Wave-SVM, we utilize the adaptive moment estimation (adam) algorithm, which confers multiple benefits, including the incorporation of adaptive learning rates, efficient memory utilization, and faster convergence during training. It is noteworthy that this paper marks the first instance of adam's application to solve an SVM model. Further, we devise an iterative algorithm to solve the optimization problems of Wave-TSVM. To empirically showcase the effectiveness of the proposed Wave-SVM and Wave-TSVM, we evaluate them on benchmark UCI and KEEL datasets (with and without feature noise) from diverse domains. Moreover, to exemplify the applicability of Wave-SVM in the biomedical domain, we evaluate it on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. The experimental outcomes unequivocally reveal the prowess of Wave-SVM and Wave-TSVM in achieving superior prediction accuracy against the baseline models. The source code

关键词： Supervised learning Pattern classification Loss function Support vector machine Twin support vector machine Wave loss function adam algorithm Alzheimer's disease

来源：评论

学校读者我要写书评

暂无评论

A new high-precision sphere-fitting method with small segment angles

引用

MEASUREMENT SCIENCE AND TECHNOLOGY 2021年第1期32卷 015012-015012页

作者： Fei, Zhigen Fu, Jixiang Ma, Jun He, Wenbin Xiao, Yanqiu Zhengzhou Univ Light Ind Henan Key Lab Intelligent Mfg Mech Equipment Zhengzhou 450002 Peoples R China

Measuring a spherical surface with small segment angle (e.g. 1 degrees) precisely is a big challenge in scientific research and engineering applications. To fit sphere surfaces with small segment angles with high precision, a five-parameter constrained nonlinear least square fitting (CNLSF) method is presented in this work, where an adam optimization algorithm with a prominent anti-interference ability is employed. We discuss various factors which influence the final uncertainty evaluation of the sphere's radius. A Monte Carlo simulation is conducted by superimposing Gaussian noise on the ideal spherical data to simulate practical instrument errors, then the relationship between the fitting bias of the sphere's radius and noise the level, the number of sampling points and the segment angle is investigated by comparing the performance of the linear least square fitting algorithm, the nonlinear least square fitting algorithm and the CNLSF algorithm. Finally, the simulated and experimental results show that the CNLSF method performs more precisely and robustly than the other algorithms, even when the corresponding segment angle approaches 1 degrees, which illustrates that it is a promising and practical method in the real measurement world.

关键词： sphere fitting constrained nonlinear least square fitting small segment angle adam algorithm

来源：评论

学校读者我要写书评

暂无评论

Data-driven product design evaluation method based on multi-stage artificial neural network

引用

APPLIED SOFT COMPUTING 2021年 103卷 107117-107117页

作者： Wang, Lei Liu, Zhengchao Wuhan Univ Technol Hubei Digital Mfg Key Lab Wuhan 430070 Peoples R China Jiangxi Univ Sci & Technol Sch Mech & Elect Engn Ganzhou 341000 Peoples R China

The application of the new generation of IT technology in the manufacturing industry has brought a rich data foundation for the research of product life cycle management. However, in the product design stage, the existing design evaluation methods have not yet effectively used the data-driven advantages, which leads to over-reliance on personal experience and inefficiency. This paper presents a new data-driven product design evaluation (PDE) method to address this problem. First, a PDE model considering multi-stage evaluation indicators of the product life cycle is established. Second, to accurately and quickly realize the nonlinear mapping between the PDE model indicators, an improved multi-stage artificial neural network (ANN) based on a hybrid optimization algorithm combining particle swarm optimization and adam (PSO-adam) is proposed. The PSO is employed for a rapid convergence during the initial stage of a global search, while adam is used to optimizing the training parameters around the global optimum adaptively. Finally, comparison experiments of the smartphone design are carried out to demonstrate the effectiveness of the multi-stage ANN and the PSO-adam. The results demonstrate the proposed approach can help designers comprehensively consider design parameters and make fast and accurate design evaluation. (C) 2021 Elsevier B.V. All rights reserved.

关键词： Data driven Product design evaluation Artificial neural network Particle swarm optimization adam algorithm

来源：评论

学校读者我要写书评

暂无评论

A Novel algorithm for Solving High-Dimensional Poisson Equations Based on Radial Basis Function Neural Networks

引用

JOURNAL OF CIRCUITS SYSTEMS AND COMPUTERS 2024年第13期33卷 2450223-2450223页

作者： Lu, Peixiao Sun, Shaoming Xinjiang Univ Sch Math & Syst Sci Urumqi 830046 Xinjiang Peoples R China Hefei Sci Isl Hefei Intelligence Hefei 230031 Anhui Peoples R China

As a widely used equation in electrostatics, the Poisson equation has significant research value in numerical solution. The basic principle of existing methods is to divide the solution domain into various grids and solve the numerical solutions at each grid node. Therefore, the accuracy of the solution is strongly correlated with the grid density divided. Based on this, this paper proposes a grid-free numerical calculation method that requires far fewer model parameters than traditional methods, and can ignore the order of the equation to solve high-dimensional Poisson equations. Given a Poisson equation, which has a certain type of boundary condition. A certain number of coordinate points are selected on the solution space and its boundary to construct a dataset. Using automatic differentiation technique to fit the differential operator in the equation, a loss function is constructed by incorporating the given boundary conditions or initial conditions, and the final numerical solution is obtained through iterative optimization algorithms. In the numerical experiment section, the algorithm proposed in this paper was used to solve the two-dimensional and three-dimensional Poisson equations with given exact solutions. The relative errors between the numerical solution and the true solution were 8.53e-4 and 6.40e-2, which are within the acceptable range. This proves that the proposed algorithm is feasible for solving the two-dimensional and three-dimensional Poisson equations with precise solutions. Secondly, the proposed algorithm is used to solve the four-dimensional Poisson equation with first-type boundary conditions, and the relative error range of the solution was within [0,0.56], which successfully extends the algorithm to solve high-dimensional Poisson equations and verifies its feasibility and efficiency in solving high-dimensional Poisson equations regardless of the dimension restriction.

关键词： Poisson equation radial basis neural (RBF) network automatic differentiation technology adam algorithm mesh free numerical algorithm

来源：评论

学校读者我要写书评

暂无评论

A general deep hybrid model for bioreactor systems: Combining first principles with deep neural networks

引用

COMPUTERS & CHEMICAL ENGINEERING 2022年 165卷

作者： Pinto, Jose Mestre, Mykaella Ramos, J. Costa, Rafael S. Striedner, Gerald Oliveira, Rui NOVA Univ Lisbon NOVA Sci & Engn Sch LAQV REQUIMTE Lisbon Portugal BOKU Univ Dept Biotechnol Vienna Austria

Numerous studies have reported the use of hybrid semiparametric systems that combine shallow neural networks with First Principles for bioprocess modeling. Here we revisit the general bioreactor hybrid model and introduce some deep learning techniques. Multi-layer networks with varying depths were combined with First Principles equations in the form of deep hybrid models. Deep learning techniques, namely the adaptive moment estimation method (adam), stochastic regularization and depth-dependent weights initialization were evaluated in a hybrid modeling context. Modified sensitivity equations are proposed for the computation of gradients in order to reduce CPU time for the training of deep hybrid models. The methods are illustrated with applications to a synthetic dataset and a pilot 50 L MUT+ Pichia pastoris process expressing a single chain antibody fragment. All in all, the results point to a systematic generalization improvement of deep hybrid models over its shallow counterpart. Moreover, the CPU cost to train the deep hybrid models is shown to be lower than for the shallow counterpart. In the pilot 50L MUT+ Pichia pastoris data set, the prediction accuracy was increased by 18.4% and the CPU decreased by 43.4%.

关键词： Hybrid semiparametric modeling Deep neural networks adam algorithm Stochastic regularization Bioprocess dynamics Pichia pastoris

来源：评论

学校读者我要写书评

暂无评论

Short-term building cooling load prediction model based on Dwdadam-ILSTM algorithm: A case study of a commercial building

引用

ENERGY AND BUILDINGS 2022年 272卷

作者： Dong, Fangnan Yu, Junqi Quan, Wei Xiang, Yalun Li, Xiang Sun, Fukang Xian Univ Architecture & Technol Xian Shaanxi Peoples R China

Traditional prediction methods of building cooling load fail to consider load changes caused by the time factor, which results in the failure of the best predicting effect and portability. To explore a more accurate cooling load predicting approach, in this paper, a new hybrid model based on decoupled weight decay adam optimized improved LSTM neural network, denoted as Dwdadam-ILSTM, is proposed. First, the cross-correlation function is used to extract features with a high degree as the input data. Then, the decoupling weight decay adaptive moment estimation (Dwdadam) optimization algorithm is utilized to optimize the parameters of the improved long short term memory (ILSTM) prediction model. Next, the ILSTM is employed to establish a prediction model. Finally, to verify the validity and prediction accu-racy of the proposed model, this paper selects the hourly cooling load data of a large commercial building in Xi'an as the research sample. Results of the experiment show that the Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Mean Square Error (MSE), Variation Coefficient of Root Mean Square Error (CV-RMSE) and Range of Relative Errors (E-R) of Dwdadam-ILSTM predicting the cooling load are 2.2%, 2.74, 5.2%, 0.061, 0.216. Compared with three version variant LSTM neural networks, the MSE and E-R values of this model are decreased by 74%-53%, 42%-49%, and 69%-64%, which has better performance in accuracy and stability. It testifies that the proposed strategy is reliable and effective. Additionally, this model can provide important guidance for daily energy management and multi -energy scheduling of buildings, which is helpful to the application of cooling load prediction. (c) 2022 Elsevier B.V. All rights reserved.

关键词： Building cooling load Short-term prediction Long short-term memory adam algorithm Predictive performance

来源：评论

学校读者我要写书评

暂无评论

CNN-based Traffic Sign Recognition

引用

计算机科学与技术汇刊（中英文版） 2022年第1期9卷 1-7页

作者： Qingkun Huang Askar Mijiti Dali Nursing Vocational College Dali 671000 Nanjing university of Finance&Economics Nanjing 210000

Background:The rapid development of the automobile industry has led to an increase in the output and holdings of automobiles year by year,which has brought huge challenges to the current traffic ***:This paper adopts a traffic sign recognition technology based on deep convolution neural network(CNN):step 1,preprocess the collected traffic sign images through gray processing and near interpolation;step 2,automatically extract image features through the convolutional layer and the pooling layer;step 3,recognize traffic signs through the fully connected layer and the Dropout ***:Artificial intelligence technology is applied to traffic management to better realize intelligent traffic assisted ***:This paper adopts an adam optimization algorithm for calculating the loss *** average accuracy of the experimental classification is 98.87%.Compared with the traditional gradient descent algorithm,the experimental model can quickly converge in a few iteration cycles.

关键词： Traffic Sign Recognition Convolution Neural Network(CNN) adam algorithm

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：