With the increment of the number of Gaussian components, the computation cost increases in the Gaussian mixture probability hypothesis density(GM-PHD) filter. Based on the theory of Chen et al, we propose an improved ...With the increment of the number of Gaussian components, the computation cost increases in the Gaussian mixture probability hypothesis density(GM-PHD) filter. Based on the theory of Chen et al, we propose an improved pruning algorithm for the GM-PHD filter, which utilizes not only the Gaussian components’ means and covariance, but their weights as a new criterion to improve the estimate accuracy of the conventional pruning algorithm for tracking very closely proximity targets. Moreover, it solves the end-less while-loop problem without the need of a second merging step. Simulation results show that this improved algorithm is easier to implement and more robust than the formal ones.展开更多
As the solutions of the least squares support vector regression machine （LS-SVRM） are not sparse, it leads to slow prediction speed and limits its applications. The defects of the ex- isting adaptive pruning algorit...As the solutions of the least squares support vector regression machine （LS-SVRM） are not sparse, it leads to slow prediction speed and limits its applications. The defects of the ex- isting adaptive pruning algorithm for LS-SVRM are that the training speed is slow, and the generalization performance is not satis- factory, especially for large scale problems. Hence an improved algorithm is proposed. In order to accelerate the training speed, the pruned data point and fast leave-one-out error are employed to validate the temporary model obtained after decremental learning. The novel objective function in the termination condition which in- volves the whole constraints generated by all training data points and three pruning strategies are employed to improve the generali- zation performance. The effectiveness of the proposed algorithm is tested on six benchmark datasets. The sparse LS-SVRM model has a faster training speed and better generalization performance.展开更多
Expert System （ES） is considered effective and efficient in agricultural production, as agricultural informationization becomes a main trend in agricultural development. ES, however, is applied unsatisfactorily in m...Expert System （ES） is considered effective and efficient in agricultural production, as agricultural informationization becomes a main trend in agricultural development. ES, however, is applied unsatisfactorily in most rural areas of China and it has considerably affected and restricted the development of the agricultural informationization. This paper proposed a solution to voice service system of ES, which was suitable for the information transmission, and it especially could help the peasants in remote regions obtain knowledge from ES through the voice service system. As for the disadvantages of massive knowledge data and slow deduction, in this system the classification method could be adopted based on the decision tree. Designing pruning algorithm to ＂trim off＂ the unrelated knowledge to the users in query course would simplify the structure of the decision tree and accelerate the speed of deduction before the inference engine deduced the knowledge required by users.展开更多
Least square support vector regression(LSSVR)is a method for function approximation,whose solutions are typically non-sparse,which limits its application especially in some occasions of fast prediction.In this paper,a...Least square support vector regression(LSSVR)is a method for function approximation,whose solutions are typically non-sparse,which limits its application especially in some occasions of fast prediction.In this paper,a sparse algorithm for adaptive pruning LSSVR algorithm based on global representative point ranking(GRPR-AP-LSSVR)is proposed.At first,the global representative point ranking(GRPR)algorithm is given,and relevant data analysis experiment is implemented which depicts the importance ranking of data points.Furthermore,the pruning strategy of removing two samples in the decremental learning procedure is designed to accelerate the training speed and ensure the sparsity.The removed data points are utilized to test the temporary learning model which ensures the regression accuracy.Finally,the proposed algorithm is verified on artificial datasets and UCI regression datasets,and experimental results indicate that,compared with several benchmark algorithms,the GRPR-AP-LSSVR algorithm has excellent sparsity and prediction speed without impairing the generalization performance.展开更多
This experiment used pumpkin Longyuanlixiang as material to research the effect of different pruning ways with one branch, two branches, and three branches on the sugar-decreasing matters during the ripeness of pumpk...This experiment used pumpkin Longyuanlixiang as material to research the effect of different pruning ways with one branch, two branches, and three branches on the sugar-decreasing matters during the ripeness of pumpkin （Cucurbita moschata） fruits. The results showed that pectin contents of pumpkin fruits during the ripeness all increased at first and then dropped in every treatment. The pectin contents under the treatments of one branch and two branches reached the high apex after pollinated for 30 d as well as for 35 d under the treatment of three branches. Pentose and chromium contents of fruits had the same trend under different pruning ways and increased during the ripenes. Three kinds of sugar-decreasing matters contents under treatment of two branches were bigger than treatment of one branch, while treatment of three branches was the lowest.展开更多
In order to improve the resource allocation mechanism of artificial immune recognition system(AIRS) and decrease the memory cells,a fuzzy logic resource allocation and memory cell pruning based AIRS(FPAIRS) is propose...In order to improve the resource allocation mechanism of artificial immune recognition system(AIRS) and decrease the memory cells,a fuzzy logic resource allocation and memory cell pruning based AIRS(FPAIRS) is proposed.In FPAIRS,the fuzzy logic is determined by a parameter,thus,the optimal fuzzy logics for different problems can be located through changing the parameter value.At the same time,the memory cells of low fitness scores are pruned to improve the classifier.This classifier was compared with other classifiers on six UCI datasets classification performance.The results show that the accuracies reached by FPAIRS are higher than or comparable to the accuracies of other classifiers,and the memory cells decrease when compared with the memory cells of AIRS.The results show that the algorithm is a high-performance classifier.展开更多
针对户外导盲场景中道路目标检测存在的复杂背景干扰及关键语义信息需求,当前目标检测算法在道路目标检测中表现出较低的准确性以及容易出现漏检的问题,为此提出一种基于YOLOv8n的道路目标检测算法OD-YOLO。基于FasterNet和SPPF构建主...针对户外导盲场景中道路目标检测存在的复杂背景干扰及关键语义信息需求,当前目标检测算法在道路目标检测中表现出较低的准确性以及容易出现漏检的问题,为此提出一种基于YOLOv8n的道路目标检测算法OD-YOLO。基于FasterNet和SPPF构建主干网络;使用FasterNet以增强特征提取能力,在SPPF模块中引入可分离大核注意力机制(large separable kernel attention,LSKA)以提高算法对道路目标整体的感知能力。提出一种新的C2f模块GAC2f,在减小模型计算量的同时提高其特征捕获能力,同时通过使用多样分支模块(diverse branch block,DBB)中结构重参数化思想优化GAC2f,在不损失模型性能的前提下,融合多种特征信息以显著提高模型精度,另一方面使用卷积门控线性单元(convolutional gated linear unit,Convolutional GLU)改进LarK中的大核卷积以优化GAC2f,使模型能够捕获更多上下文信息。提出一种轻量级非对称检测头PADH,在提高模型性能的同时减少参数量,并使用PIoUv2改进原有的损失函数,通过基于层自适应稀疏度的量级剪枝(layer-adaptive sparsity for the magnitude-based pruning,LAMP)操作进一步优化算法模型。实验结果表明,在公共人行道路目标数据集WOTR上,OD-YOLO与YOLOv8n相比,经过剪枝后模型参数量同为3×10^(6),但mAP@0.5、mAP@0.5:0.95分别提升3.4和4.1个百分点,证明算法OD-YOLO在面向户外导盲场景的道路目标检测中可以达到预期的效果。展开更多
基金supported by the National Natural Science Foundation of China(61703228)
文摘With the increment of the number of Gaussian components, the computation cost increases in the Gaussian mixture probability hypothesis density(GM-PHD) filter. Based on the theory of Chen et al, we propose an improved pruning algorithm for the GM-PHD filter, which utilizes not only the Gaussian components’ means and covariance, but their weights as a new criterion to improve the estimate accuracy of the conventional pruning algorithm for tracking very closely proximity targets. Moreover, it solves the end-less while-loop problem without the need of a second merging step. Simulation results show that this improved algorithm is easier to implement and more robust than the formal ones.
基金supported by the National Natural Science Foundation of China (61074127)
文摘As the solutions of the least squares support vector regression machine （LS-SVRM） are not sparse, it leads to slow prediction speed and limits its applications. The defects of the ex- isting adaptive pruning algorithm for LS-SVRM are that the training speed is slow, and the generalization performance is not satis- factory, especially for large scale problems. Hence an improved algorithm is proposed. In order to accelerate the training speed, the pruned data point and fast leave-one-out error are employed to validate the temporary model obtained after decremental learning. The novel objective function in the termination condition which in- volves the whole constraints generated by all training data points and three pruning strategies are employed to improve the generali- zation performance. The effectiveness of the proposed algorithm is tested on six benchmark datasets. The sparse LS-SVRM model has a faster training speed and better generalization performance.
基金Supported by Northeast Agricultural University Doctoral Development FoundationChina Postdoctoral Science Foundation
文摘Expert System （ES） is considered effective and efficient in agricultural production, as agricultural informationization becomes a main trend in agricultural development. ES, however, is applied unsatisfactorily in most rural areas of China and it has considerably affected and restricted the development of the agricultural informationization. This paper proposed a solution to voice service system of ES, which was suitable for the information transmission, and it especially could help the peasants in remote regions obtain knowledge from ES through the voice service system. As for the disadvantages of massive knowledge data and slow deduction, in this system the classification method could be adopted based on the decision tree. Designing pruning algorithm to ＂trim off＂ the unrelated knowledge to the users in query course would simplify the structure of the decision tree and accelerate the speed of deduction before the inference engine deduced the knowledge required by users.
基金supported by the Science and Technology on Space Intelligent Control Laboratory for National Defense(KGJZDSYS-2018-08)。
文摘Least square support vector regression(LSSVR)is a method for function approximation,whose solutions are typically non-sparse,which limits its application especially in some occasions of fast prediction.In this paper,a sparse algorithm for adaptive pruning LSSVR algorithm based on global representative point ranking(GRPR-AP-LSSVR)is proposed.At first,the global representative point ranking(GRPR)algorithm is given,and relevant data analysis experiment is implemented which depicts the importance ranking of data points.Furthermore,the pruning strategy of removing two samples in the decremental learning procedure is designed to accelerate the training speed and ensure the sparsity.The removed data points are utilized to test the temporary learning model which ensures the regression accuracy.Finally,the proposed algorithm is verified on artificial datasets and UCI regression datasets,and experimental results indicate that,compared with several benchmark algorithms,the GRPR-AP-LSSVR algorithm has excellent sparsity and prediction speed without impairing the generalization performance.
基金Supported by Heilongjiang Province Natural Science Foundation (C2004-24)
文摘This experiment used pumpkin Longyuanlixiang as material to research the effect of different pruning ways with one branch, two branches, and three branches on the sugar-decreasing matters during the ripeness of pumpkin （Cucurbita moschata） fruits. The results showed that pectin contents of pumpkin fruits during the ripeness all increased at first and then dropped in every treatment. The pectin contents under the treatments of one branch and two branches reached the high apex after pollinated for 30 d as well as for 35 d under the treatment of three branches. Pentose and chromium contents of fruits had the same trend under different pruning ways and increased during the ripenes. Three kinds of sugar-decreasing matters contents under treatment of two branches were bigger than treatment of one branch, while treatment of three branches was the lowest.
基金Project(61170199)supported by the National Natural Science Foundation of ChinaProject(11A004)support by the Scientific Research Fund of Education Department of Hunan Province,China
文摘In order to improve the resource allocation mechanism of artificial immune recognition system(AIRS) and decrease the memory cells,a fuzzy logic resource allocation and memory cell pruning based AIRS(FPAIRS) is proposed.In FPAIRS,the fuzzy logic is determined by a parameter,thus,the optimal fuzzy logics for different problems can be located through changing the parameter value.At the same time,the memory cells of low fitness scores are pruned to improve the classifier.This classifier was compared with other classifiers on six UCI datasets classification performance.The results show that the accuracies reached by FPAIRS are higher than or comparable to the accuracies of other classifiers,and the memory cells decrease when compared with the memory cells of AIRS.The results show that the algorithm is a high-performance classifier.
文摘针对户外导盲场景中道路目标检测存在的复杂背景干扰及关键语义信息需求,当前目标检测算法在道路目标检测中表现出较低的准确性以及容易出现漏检的问题,为此提出一种基于YOLOv8n的道路目标检测算法OD-YOLO。基于FasterNet和SPPF构建主干网络;使用FasterNet以增强特征提取能力,在SPPF模块中引入可分离大核注意力机制(large separable kernel attention,LSKA)以提高算法对道路目标整体的感知能力。提出一种新的C2f模块GAC2f,在减小模型计算量的同时提高其特征捕获能力,同时通过使用多样分支模块(diverse branch block,DBB)中结构重参数化思想优化GAC2f,在不损失模型性能的前提下,融合多种特征信息以显著提高模型精度,另一方面使用卷积门控线性单元(convolutional gated linear unit,Convolutional GLU)改进LarK中的大核卷积以优化GAC2f,使模型能够捕获更多上下文信息。提出一种轻量级非对称检测头PADH,在提高模型性能的同时减少参数量,并使用PIoUv2改进原有的损失函数,通过基于层自适应稀疏度的量级剪枝(layer-adaptive sparsity for the magnitude-based pruning,LAMP)操作进一步优化算法模型。实验结果表明,在公共人行道路目标数据集WOTR上,OD-YOLO与YOLOv8n相比,经过剪枝后模型参数量同为3×10^(6),但mAP@0.5、mAP@0.5:0.95分别提升3.4和4.1个百分点,证明算法OD-YOLO在面向户外导盲场景的道路目标检测中可以达到预期的效果。
