检索结果-内蒙古大学图书馆

IEEE International Conference on Image Processing

作者： Sachin S. Talathi Qualcomm Research Center

ISBN: (纸本)9781479983407

Recently sequential model based optimization (SMBO) has emerged as a promising hyper-parameter optimization strategy in machine learning. In this work, we investigate SMBO to identify architecture hyper-parameters of deep convolution networks (DCNs) object recognition. We propose a simple SMBO strategy that starts from a set of random initial DCN architectures to generate new architectures, which on training perform well on a given dataset. Using the proposed SMBO strategy we are able to identify a number of DCN architectures that produce results that are comparable to state-of-the-art results on object recognition benchmarks.

关键词： hyper-parameter optimization Deep convolution networks Sequential model based optimization object recognition parameter optimization ARCHITECTURE Network new buildings hyperparameter WWOX gene

来源：评论

学校读者我要写书评

暂无评论

Surrogate network-based sparseness hyper-parameter optimization for deep expression recognition

引用

PATTERN RECOGNITION 2021年 111卷 107701-107701页

作者： Xie, Weicheng Chen, Wenting Shen, Linlin Duan, Jinming Yang, Meng Shenzhen Univ Sch Comp Sci & Software Engn Shenzhen 518060 Peoples R China Shenzhen Inst Artificial Intelligence & Robot Soc Shenzhen Peoples R China Shenzhen Univ Guangdong Key Lab Intelligent Informat Proc Shenzhen Peoples R China Univ Birmingham Sch Comp Sci Birmingham W Midlands England Sun Yat Sen Univ Sch Data & Comp Sci Guangzhou Guangdong Peoples R China

For facial expression recognition, the sparseness constraints of the features or weights can improve the generalization ability of a deep network. However, the optimization of the hyper-parameters in fusing different sparseness strategies demands much computation, when the traditional gradient-based algorithms are used. In this work, an iterative framework with surrogate network is proposed for the optimization of hyper-parameters in fusing different sparseness strategies. In each iteration, a network with significantly smaller model complexity is fitted to the original large network based on four Euclidean losses, where the hyper-parameters are optimized with heuristic optimizers. Since the surrogate network uses the same deep metrics and embeds the same hyper-parameters as the original network, the optimized hyper-parameters are then used for the training of the original deep network in the next iteration. While the performance of the proposed algorithm is justified with a tiny model, i.e. LeNet on the FER2013 database, our approach achieved competitive performances on six publicly available expression datasets, i.e., FER2013, CK+, Oulu-CASIA, MMI, AFEW and AffectNet. (C) 2020 Elsevier Ltd. All rights reserved.

关键词： Expression recognition Deep sparseness strategies hyper-parameter optimization Surrogate network Heuristic optimizer

来源：评论

学校读者我要写书评

暂无评论

An efficient hyper-parameter optimization method for supervised learning

引用

APPLIED SOFT COMPUTING 2022年 126卷

作者： Shi, Ying Qi, Hui Qi, Xiaobo Mu, Xiaofang Taiyuan Normal Univ Sch Comp Sci & Technol Jinzhong 030619 Shanxi Peoples R China Shanxi Univ Sch Comp & Informat Technol Taiyuan 030006 Shanxi Peoples R China

Supervised learning is an important tool for data mining and knowledge discovery. The hyper -parameter in learning models usually has a significant impact on the generalization performance of supervised learning model. Although some state-of-the-art hyper-parameter optimizers, such as cross-validation (CV) and its improvements, have been widely used in many applications, there still exist some limitations including the low efficiency and the error variation from data partition. To address these issues, we propose the sign similarity to distinguish between over-fitting and under -fitting in supervised learning. On this basis, the minimal symmetric similarity criterion (MSSC) is proposed to optimize hyper-parameters. It provides an equivalent condition of well-fitting, i.e., the well-fitted hyper-parameter should have a minimal symmetric similarity. Compared with CV, the criterion is more efficient and could avoid error variation as the symmetric similarity is based on the whole data set without partition. The reasonableness of the proposed MSSC is proved for well-posed learning problems. Also, corresponding assumptions are explained by means of theoretical analysis and empirical verification. Both theoretical analysis and experimental results indicate that the efficiency of MSSC is significantly higher than popular optimizers like 10-CV. Experimental results demonstrate that MSSC is comparable with or outperforms the state-of-the-art optimizers in three kinds of learning tasks from the statistical perspective. (C) 2022 Elsevier B.V. All rights reserved.

关键词： Supervised learning hyper-parameter optimization Cross-validation Sign similarity Minimal symmetric similarity criterion

来源：评论

学校读者我要写书评

暂无评论

Multi-objective simulated annealing for hyper-parameter optimization in convolutional neural networks

引用

PEERJ COMPUTER SCIENCE 2021年 7卷 e338页

作者： Gulcu, Ayla Kus, Zeki Fatih Sultan Mehmet Univ Comp Sci Istanbul Turkey

In this study, we model a CNN hyper-parameter optimization problem as a bi-criteria optimization problem, where the first objective being the classification accuracy and the second objective being the computational complexity which is measured in terms of the number of floating point operations. For this bi-criteria optimization problem, we develop a Multi-Objective Simulated Annealing (MOSA) algorithm for obtaining high-quality solutions in terms of both objectives. CIFAR-10 is selected as the benchmark dataset, and the MOSA trade-off fronts obtained for this dataset are compared to the fronts generated by a single-objective Simulated Annealing (SA) algorithm with respect to several front evaluation metrics such as generational distance, spacing and spread. The comparison results suggest that the MOSA algorithm is able to search the objective space more effectively than the SA method. For each of these methods, some front solutions are selected for longer training in order to see their actual performance on the original test set. Again, the results state that the MOSA performs better than the SA under multi-objective setting. The performance of the MOSA configurations are also compared to other search generated and human designed state-of-the-art architectures. It is shown that the network configurations generated by the MOSA are not dominated by those architectures, and the proposed method can be of great use when the computational complexity is as important as the test accuracy.

关键词： Multi-objective Simulated annealing Convolutional neural networks hyper-parameter optimization

来源：评论

学校读者我要写书评

暂无评论

Grid search with a weighted error function: hyper-parameter optimization for financial time series forecasting

引用

APPLIED SOFT COMPUTING 2024年 154卷

作者： Zhao, Yuan Zhang, Weiguo Liu, Xiufeng South China Univ Technol Sch Business Adm Guangzhou 510641 Peoples R China Guangzhou Financial Serv Innovat & Risk Management Guangzhou 510641 Peoples R China Shenzhen Univ Coll Management Shenzhen 518060 Peoples R China Tech Univ Denmark Dept Technol Management & Econ DK-2800 Lyngby Denmark

Financial time series forecasting is a difficult task due to the complexity and volatility of financial markets. Machine learning models have been applied to tackle this task, but finding their optimal hyper -parameters with less time and ensuring the prediction accuracy of models are still significant challenges. Existing methods such as GridSearch with cross -validation (GridsearchCV) for optimizing the hyper -parameters are time-consuming for complex models or large search spaces, and they do not ensure that the model has excellent predictive accuracy. To address these challenges, we propose a novel method called GridsearchWEF that uses grid search with a weighted error function. This method aims to reduce the time cost of hyperparameter optimization for machine learning models and guarantee their prediction performance. We conduct an empirical analysis of crude oil return forecasting using four machine learning models: RF, GBDT, SVR, and LASSO. We compare the performance of these models using GridsearchCV, random search with cross -validation (RandomizedSearchCV), Bayes optimization with cross -validation (BayesSearchCV), and GridsearchWEF. The results show that GridsearchWEF outperforms the other methods in terms of hyper -parameter optimization, modeling efficiency, prediction accuracy, and economic values. In particular, the time of all models using GridSearchWEF is less than 30 s, which is much less than other algorithms. GridsearchWEF is a more efficient and superior method for hyper -parameter optimization in financial time series forecasting.

关键词： Time series forecasting Machine learning hyper-parameter optimization Modeling efficiency Economic value

来源：评论

学校读者我要写书评

暂无评论

Incremental-based YoloV3 model with hyper-parameter optimization for product image classification in E-commerce sector

引用

APPLIED SOFT COMPUTING 2024年 165卷

作者： Dutta, Munmi Ganguly, Amrita Assam Engn Coll Dept Elect Engn Gauhati 781013 Assam India

Over the past few years, the E-commerce industry has grown tremendously for selling products to consumers. Here, the consumer can easily purchase the products from their residing seats and gets the products at the door step. Also, the image quality is suggested and also the image classification is intruded to meet the consumer prerequisite. Most of the image classification models are developed by machine learning approaches to improve the reliability, scalability, and accuracy level. Due to the heterogeneous nature of images, it restricts the features and lengthy dimensions for datasets which it degrades the performance whereas the classification of products becomes a cumbersome situation. In order to resolve the problem raised by existing methods, a novel product image classification model is proposed for the fashion E-commerce sector. The prime aspect of the proposed model is to build an effective E-commerce sector for a layman, where every individual can purchase the products appropriately. Initially, the dataset is constructed with diverse product images that are obtained from different Ecommerce fields with various classes. Consequently, a novel Incremental-based Improved YoloV3 with hyperparameter optimization (II-Yolov3-HO) model is introduced to enhance the classification performance. In the proposed model, some of the hyper parameters like convolution and pooling layers are replaced with the Visual Geometry Group16 (VGG16) model, the remaining convolutional layers are enhanced by the atrous convolution operation, and pooling layers are modified with Atrous Spatial Pyramid Pooling (ASPP). In entire architecture, the weight factor is optimized using Predator-based Squirrel Search Algorithm (PDP-SSA) algorithm. Thus, the new E-commerce sector becomes free to categorize the products, and if it deals with a new product, it is declared as unknown data, which is then trained again in the proposed model. Hence, the concept of classification score matching with the

关键词： Product image classification Fashion E -commerce sector Unknown classes hyper-parameter optimization Predator-based Squirrel Search Algorithm

来源：评论

学校读者我要写书评

暂无评论

Application of a Stochastic Schemata Exploiter for Multi-Objective hyper-parameter optimization of Machine Learning

引用

REVIEW OF SOCIONETWORK STRATEGIES 2023年第2期17卷 179-213页

作者： Makino, Hiroya Kita, Eisuke Nagoya Univ Grad Sch Informat Nagoya Japan

The Stochastic Schemata Exploiter (SSE), one of the Evolutionary Algorithms, is designed to find the optimal solution of a function. SSE extracts common schemata from individual sets with high fitness and generates individuals from the common schemata. For hyper-parameter optimization, the initialization method, the schema extraction method, and the new individual generation method, which are characteristic processes in SSE, are extended. In this paper, an SSE-based multi-objective optimization for AutoML is proposed. AutoML gives good results in terms of model accuracy. However, if only model accuracy is considered, the model may be too complex. Such complex models cannot always be allowed because of the long computation time. The proposed method maximizes the stacking model accuracy and minimizes the model complexity simultaneously. When compared with existing methods, SSE has interesting features such as fewer control parameters and faster convergence properties. The visualization method makes the optimization process transparent and helps users understand the process.

关键词： AutoML Stochastic schemata exploiter Evolutionary algorithm hyper-parameter optimization

来源：评论

学校读者我要写书评

暂无评论

Research about pruning hyper-parameter optimization method based on transfer learning in geographic information system

引用

ARABIAN JOURNAL OF GEOSCIENCES 2021年第5期14卷 404-404页

作者： Zhang, Xiaohang Li, Yuqi Li, Zhengren Beijing Univ Posts & Telecommun Sch Modern Post Beijing 100876 Peoples R China Beijing Univ Posts & Telecommun Sch Econ & Management Beijing 100876 Peoples R China

Recently, researchers have found that some automated optimization methods and techniques can speed up the whole search process and can obtain better model hyper-parameter configurations. In this paper, the optimized initial hyper-parameter configuration is obtained through the loss function defined under the definition of meta-features combined with the algorithm of transfer learning. Furthermore, the sequence model-based hyper-parameter optimization model (SMBO model) is transformed by introducing a pruning strategy. By categorizing the hyper-parameters according to their importance, the non-important hyper-parameters are eliminated but important hyper-parameters are strengthened, so the final hyper-parameter optimization configuration is obtained. It is the parameter optimization design based on this secondary optimization that makes it possible to quickly search and select relevant information and carry out data application. The main goal of this paper is to clarify the mathematical principle of secondary optimization and combine it with geographic information system to create the value development of geographic information database.

关键词： hyper-parameter optimization Automated machine learning algorithm Transfer learning Pruning strategy SMBO model Geographic information system (GIS)

来源：评论

学校读者我要写书评

暂无评论

A Novel Method Based on Line-Segment Visualizations for hyper-parameter optimization in Deep Networks

引用

INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE 2018年第3期32卷 1851002-1851002页

作者： Tang, Xue-song Ding, Yongsheng Hao, Kuangrong Donghua Univ Dept Informat Sci Shanghai 201620 Peoples R China

Recently, deep learning has been widely applied in various areas and achieved remarkable research findings. The major reason that makes the deep learning paradigm successful is that it can effectively learn a hierarchical feature structure for the training data. However, most deep learning algorithms rely on massive well-labeled training datasets and hyper-parameter configurations. This paper proposed a novel methodology that uses the geometric characteristics of line-segment representations to optimize the hyper-parameters for the deep networks. The methodology is applied to a line-segment-based stacked auto-encoder to verify its effectiveness. It is found that the line-segment-based visualizations can increase the interpretability of the deep models and facilitate the configurations for the hyper-parameters.

关键词： hyper-parameter optimization line-segment representation feature visualizations deep learning

来源：评论

学校读者我要写书评

暂无评论

Federated learning with hyper-parameter optimization

引用

JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES 2023年第9期35卷

作者： Kundroo, Majid Kim, Taehong Chungbuk Natl Univ Sch Informat & Commun Engn Cheongju 28644 South Korea

Federated Learning is a new approach for distributed training of a deep learning model on data scattered across a large number of clients while ensuring data privacy. However, this approach faces certain limitations, including a longer convergence time compared to typical deep learning models. Existing federated optimization algorithms often employ the same hyper-parameters for all clients, disregarding potential system heterogeneity and varying local data availability, which contributes to an even longer convergence time and more communication rounds. To address this challenge, we propose FedHPO, a new federated optimization algorithm that adaptively modifies hyper-parameters of each client's local model during training, such as learning rate and epochs. This adaptability facilitates quicker convergence of the client's local model, which in turn helps the global model converge faster, consequently reducing overall convergence time and the required communication rounds. In addition, FedHPO does not require any additional complexity since each client adjusts hyper-parameters independently based on the training results obtained in each epoch. In our evaluation, we compare FedHPO with other algorithms, such as FedAVG, FedAVGM, FedProx, and FedYogi, using both IID and non-IID distributed datasets. The results demonstrate the promising outcomes of FedHPO, showcasing reduced convergence time and fewer required communication rounds in comparison to alternative algorithms. (c) 2023 The Author(s). Published by Elsevier B.V. on behalf of King Saud University. This is an open access article under the CC BY-NC-ND license (http:***-nc-ndj4.0j).

关键词： Federated learning hyper-parameter optimization Convergence time Communication cost

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：