检索结果-内蒙古大学图书馆

Meta-Heuristic Search optimization and its application to Time Series Forecasting Model

INTELLIGENT SYSTEMS WITH APPLICATIONS 2022年 16卷

作者： Khairalla, Mergani Nile Valley Univ Fac Comp Sci & Informat Technol Atbara 346 Sudan

Recently, ensemble learning has been widely used to generate better generalization ability in time series forecasting. The estimation accuracy of the ensemble model depends on the hyper-parameters tuning. The purpose of this research is to introduce a novel kernel-based ensemble machines (KBEM) to determine the best kernel structure with fine-tune hyper-parameters. The optimization method using a hybrid scheme is a combination of meta-heuristic search and local weight learning (LWL) in the process of generating ensemble layers. In order to determine the efficiency of KBEM method support vector regression (SVR) is utilized as a base learner for ensemble. The case studies are based on six benchmarks and global oil consumption data sets. The final output of the proposed model is compared with the current state-of-the-arts tuning techniques, which show that KBEM has achieved obvious advantages in terms of time complexity and performance analysis. Therefore, the KBEM scheme can be regarded as a promising method for automatic kernel regularization and hyper-parameter adjustment of integrated models.

关键词： Time series forecasting Ensemble learning Ensemble tuning hyper-parameters optimization Meta-heuristic Search

来源：评论

学校读者我要写书评

暂无评论

MODE-CNN: A fast converging multi-objective optimization algorithm for CNN-based models

引用

APPLIED SOFT COMPUTING 2021年 109卷 107582-107582页

作者： Inik, Ozkan Altiok, Mustafa Ulker, Erkan Kocer, Baris Tokat Gaziosmanpasa Univ Dept Comp Engn Tasliciftlik Campus TR-60250 Tokat Turkey Konya Tech Univ Dept Comp Engn Alaaddin Keykubat Campus TR-42250 Konya Turkey

Convolutional neural networks (CNNs) have been used to solve many problems in computer science with a high level of success, and have been applied in many fields in recent years. However, most of the designs of these models are still tuned manually;obtaining the highest performing CNN model is therefore very time-consuming, and is sometimes not achievable. Recently, researchers have started using optimization algorithms for the automatic adjustment of the hyper-parameters of CNNs. In particular, single-objective optimization algorithms have been used to achieve the highest network accuracy for the design of a CNN. When these studies are examined, it can be seen that the most significant problem in the optimization of the parameters of a CNN is that a great deal of time is required for tuning. Hence, optimization algorithms with high convergence rates are needed for the parameter optimization of deep networks. In this study, we first develop an algorithm called MODE-CNN, based on the multi-objective differential evolution (MODE) algorithm for parameter optimization of CNN or CNN-based methods. MODE-CNN is then compared with four different multi-objective optimization algorithms. This comparison is carried out using 16 benchmark functions and four different metrics, with 100 independent runs. It is observed that the algorithm is robust and competitive compared to alternative approaches, in terms of its accuracy and convergence. Secondly, the MODE-CNN algorithm is used in the parameter optimization of a CNN-based method, developed previously by the authors, for the segmentation and classification of medical images. In this method, there are three parameters that influence the test time and accuracy: the general stride (GS), neighbour distance (ND), and patch accuracy (PA). These parameters need to be optimized to give the highest possible accuracy and lowest possible test time. With the MODE-CNN algorithm, the most appropriate GS, ND, and PA values are obtained for the

关键词： Deep learning CNN hyper-parameters optimization Multi-objective MODE

来源：评论

学校读者我要写书评

暂无评论

Novel Artificial Immune Networks-based optimization of shallow machine learning (ML) classifiers

引用

EXPERT SYSTEMS WITH APPLICATIONS 2021年 165卷 113834-113834页

作者： Kanwal, Summrina Hussain, Amir Huang, Kaizhu Saudi Elect Univ Dept Comp Sci Riyadh Saudi Arabia Edinburgh Napier Univ Edinburgh City Of Edinburgh Scotland Xian Jiaotong Liverpool Univ Dept Elect & Elect Engn Suzhou 215123 Peoples R China

Artificial Immune Networks (AIN) is a population-based evolutionary algorithm that is inspired by theoretical immunology. It applies ideas and metaphors from the biological immune system to solve multi-disciplinary problems. This paper presents a novel application of the AIN for optimizing shallow machine learning (ML) classification algorithms. AIN accomplishes this task by searching the best hyper-parameter set for a specific classification algorithm (also termed model selection), which minimizes training error and enhances the generalization capability of the algorithm. We present a convergence analysis of the proposed algorithm and employ it in conjunction with selected, well-known ML classifiers, namely, an extreme learning machine (ELM), a support vector machine (SVM) and an echo state network (ESN). The performance is evaluated in terms of classification accuracy and learning time, using a range of benchmark datasets, and compared against grid search as well as evolutionary strategy (ES)-based optimization techniques. An empirical study with different datasets demonstrates improved classification accuracy of SVM, from 2% to 5%, for ESN from 3% to 6%, whereas in the case of ELM from 3% to 9%. Comparative simulation results demonstrate the potential of AIN as an alternative optimizer for shallow ML algorithms.

关键词： Artificial Immune Network (AIN) Echo State Network (ESN) Extreme Learning Machine (ELM) Support Vector Machine (SVM) hyper-parameters optimization

来源：评论

学校读者我要写书评

暂无评论

Deep Forest Regression Based on Dynamic State Transition optimization Algorithm

Deep Forest Regression Based on Dynamic State Transition Opt...

引用

Chinese Automation Congress (CAC)

作者： Xia, Heng Tang, Jian Qiao, Junfei Beijing Univ Technol Fac Informat Technol Beijing Peoples R China

ISBN: (纸本)9781728176871

As a deep algorithm of non-neural network structure, deep forest regression (DFR) can be used to build soft measuring models of difficult-to-measure key parameters. However, as a kind of deep learning, the optimization of hyperparameters has become an inevitable problem in DFR. To solve above problem, an improved dynamic state transition algorithm (DSTA) is used to optimize the hyper-parameters of the model. To achieved more accurate optimization process, the error change rate is used to fine-tuning the state factor during the iteration process, which is further improved with gradient-based refinement. Finally, simulation experiments are performed on the benchmark data set, and satisfactory simulation results show the effectiveness of the proposed approach.

关键词： deep forest regression hyper-parameters optimization dynamic state transition algorithm

来源：评论

学校读者我要写书评

暂无评论

The Image Classification Method with CNN-XGBoost Model Based on Adaptive Particle Swarm optimization

引用

INFORMATION 2021年第4期12卷 156页

作者： Jiao, Wenjiang Hao, Xingwei Qin, Chao Shandong Univ Sch Software Jinan 250101 Peoples R China Shandong Univ Finance & Econ Sch Comp Sci & Technol Jinan 250014 Peoples R China

CNN is particularly effective in extracting spatial features. However, the single-layer classifier constructed by activation function in CNN is easily interfered by image noise, resulting in reduced classification accuracy. To solve the problem, the advanced ensemble model XGBoost is used to overcome the deficiency of a single classifier to classify image features. To further distinguish the extracted image features, a CNN-XGBoost image classification model optimized by APSO is proposed, where APSO optimizes the hyper-parameters on the overall architecture to promote the fusion of the two-stage model. The model is mainly composed of two parts: feature extractor CNN, which is used to automatically extract spatial features from images;feature classifier XGBoost is applied to classify features extracted after convolution. In the process of parameter optimization, to overcome the shortcoming that traditional PSO algorithm easily falls into a local optimal, the improved APSO guide the particles to search for optimization in space by two different strategies, which improves the diversity of particle population and prevents the algorithm from becoming trapped in local optima. The results on the image set show that the proposed model gets better results in image classification. Moreover, the APSO-XGBoost model performs well on the credit data, which indicates that the model has a good ability of credit scoring.

关键词： figure classification binary classification hyper-parameters optimization credit scoring

来源：评论

学校读者我要写书评

暂无评论

Early diagnosis of Parkinson's disease from multiple voice recordings by simultaneous sample and feature selection

引用

EXPERT SYSTEMS WITH APPLICATIONS 2019年 137卷 22-28页

作者： Ali, Liaqat Zhu, Ce Zhou, Mingyi Liu, Yipeng UESTC Sch Informat & Commun Engn Chengdu Sichuan Peoples R China Univ Sci & Technol Dept Elect Engn Bannu Pakistan

Parkinson's disease (PD) is a serious neurodegenerative disorder. It is reported that more than 90% of PD patients have voice impairments. Multiple types of voice recordings have been used for PD detection. Previous work indicates that the use of multiple types of samples per subject degenerates PD detection accuracy. In this paper, we validate it, and propose a two dimensional data selection method for sample and feature selection. The proposed method ranks features by using chi-square statistical model, searches optimal subset of the ranked features and iteratively selects samples. Experimental results show that the proposed method outperforms the state-of-the-art methods in terms of PD detection accuracy on multiple types of voice data. (C) 2019 Elsevier Ltd. All rights reserved.

关键词： Deep neural network Feature selection hyper-parameters optimization Parkinson'S disease Sample selection

来源：评论

学校读者我要写书评

暂无评论

Fault Diagnosis of Rolling Bearing Based on SDAE and PSO-DBN 31

Fault Diagnosis of Rolling Bearing Based on SDAE and PSO-DBN

引用

31st Chinese Control And Decision Conference (CCDC)

作者： Wang, Zhihao Sun, Teng Tian, Xincheng Shandong Univ Sch Control Sci & Engn Jinan 250014 Peoples R China Weichai Power Co Ltd Business Management & Informat Technol Wei Fang 261000 Peoples R China

ISBN: (纸本)9781728101057

A new fault diagnosis method for rolling bearing based on two-step cascaded system with two deep network is presented in this paper. In view of the low accuracy of traditional diagnosis algorithm, Stacked Denoising Auto-Encoder (SDAE) model as the first network is used to extract the basic and shallow feature of the fault signal;in order to acquire more robust and deep feature representation, Deep Belief Network (DBN) is configured as the second network. However, as for specific fault diagnosis problems, the number of hidden layer nodes, learning rate and momentum factor will directly affect the diagnosis result of DBN model. Therefore, this paper adopts particle swarm optimization (PSO) algorithm to adaptively select the hyper-parameters of DBN to determine the optimal structure of network, finally realizes the classification of multiple faults. Rolling bearing fault simulation and experiments have been conducted under single load condition to verify the effectiveness of the proposed algorithm. Experimental results obviously demonstrate that. from the aspects of generalization capability and classification performance, this algorithm is superior to support vector machine (SVM), back propagation neural network (BPNN) and grey relational analysis (GRA).

关键词： Fault Diagnosis Stacked Denoising Auto-Encoder Deep Belief Network hyper-parameters optimization

来源：评论

学校读者我要写书评

暂无评论

A Note on Using Performance and Data Profiles for Training Algorithms

引用

ACM TRANSACTIONS ON MATHEMATICAL SOFTWARE 2019年第2期45卷 20-20页

作者： Porcelli, March Erita Toint, Philippe L. Univ Firenze Dipartimento Ingn Ind Viale GB Morgagni 40 Florence Italy Univ Namur Namur Ctr Complex Syst NAXYS 61 Rue Bruxelles B-5000 Namur Belgium

This article shows how to use performance and data profile benchmarking tools to improve the performance of algorithms. We propose to achieve this goal by defining and approximately solving suitable optimization problems involving the parameters of the algorithm under consideration. Because these problems do not have derivatives and may involve integer variables, we suggest using a mixed-integer derivative-free optimizer for this task. A numerical illustration is presented (using the BFO package), which indicates that the obtained gains are potentially significant.

关键词： Algorithmic design hyper-parameters optimization trainable codes derivative-free optimization

来源：评论

学校读者我要写书评

暂无评论

DeepDIVA: A Highly-Functional Python Framework for Reproducible Experiments 16

DeepDIVA: A Highly-Functional Python Framework for Reproduci...

引用

16th International Conference on Frontiers in Handwriting Recognition (ICFHR)

作者： Alberti, Michele Pondenkandath, Vinaychandran Wursch, Marcel Ingold, Rolf Liwicki, Marcus Univ Fribourg Document Image & Voice Anal Grp DIVA Fribourg Switzerland Lulea Univ Technol Machine Learning Grp Lulea Sweden

ISBN: (纸本)9781538658758

We introduce DeepDIVA: an infrastructure designed to enable quick and intuitive setup of reproducible experiments with a large range of useful analysis functionality. Reproducing scientific results can be a frustrating experience, not only in document image analysis but in machine learning in general. Using DeepDIVA a researcher can either reproduce a given experiment or share their own experiments with others. Moreover, the framework offers a large range of functions, such as boilerplate code, keeping track of experiments, hyper-parameter optimization, and visualization of data and results. To demonstrate the effectiveness of this framework, this paper presents case studies in the area of handwritten document analysis where researchers benefit from the integrated functionality. DeepDIVA is implemented in Python and uses the deep learning framework PyTorch. It is completely open source(1), and accessible as Web Service through DIVAServices(2).

关键词： Framework Open-Source Deep Learning Neural Networks Reproducible Research Machine Learning hyper-parameters optimization Python

来源：评论

学校读者我要写书评

暂无评论

LS-SVM-Based Prediction Model of Tread Wear Optimized by PSO-GA-LM

LS-SVM-Based Prediction Model of Tread Wear Optimized by PSO...

引用

2nd International Conference on Cloud Computing and Security (ICCCS)

作者： Hua, Sha Yuan, Jiabin Ding, Weijie Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing Jiangsu Peoples R China

ISBN: (纸本)9783319486741;9783319486734

The wheel wear is a dynamic phenomenon that varies with many mechanical and geometrical factors. Accurately estimating wheel wear is a vital issue in wheel maintance. This paper presents a nature-inspired metaheuristic regression method for precisely predicting wheel status that combines least squares support vector machine (LS-SVM) with a novel PSO-GA-LM algorithm. The PSO-GA-LM algorithm integrates Particle Swarm optimization (PSO), Genetic Algorithm (GA) and Logistic Map (LM). The method is used to optimize the hyper-parameters of the LS-SVM model. The proposed model was constructed with datasets of the tread wear derived from Taiyuan North Locomotive Depot. Analytical results show that the novel optimized prediction model is superior to others in predicting tread wear with lower RMSE (0.037MPa), MAE (0.027MPa) and MAPE (0.0008 %).

关键词： LS-SVM hyper-parameters optimization PSO-GA-LM algorithm Tread wear

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：