本研究旨在探索基于测光图像的特殊恒星搜寻方法，以提高对热亚矮星等特殊恒星的识别效率。特殊恒星作为恒星物理学的关键对象，对理解宇宙的诞生与进化具有重要意义。本项目利用深度学习目标检测算法YOLOv5，结合机器学习和人工智能技术，对斯隆数字化巡天（SDSS）项目中的测光图像进行分析，以实现对特殊恒星的高效搜寻。研究团队首先对SDSS数据进行交叉处理，获取了3457条有效数据，随后通过数据增强和标注，构建了适合YOLOv5模型训练的数据集。实验结果表明，改进后的YOLOv5m模型在热亚矮星检测方面表现出较高的Precision、Recall和mAP@0.5，验证了该方法在特殊恒星搜寻中的有效性和应用潜力。 (2024-06-10, Others, 0KB, 下载0次)
http://www.pudn.com/Download/item/id/1718010780209825.html
农牧行业专业定制 种、养、加、销、运智慧高效一体化解决方案。主要应用在繁育、养殖、社会化服务、屠 宰加工、金融保险、销售、运营、监管等畜牧业全产业链的各重要环节,是为农牧民、 合作社、企业等提供饲养技术、繁育服务、远程诊断、农资流通、金融保险、产品销 售、市场分析、运营监管等服务的服务型大数据云计算平台。
Specialized customized intelligent and efficient integrated solutions for planting, breeding, processing, marketing and operation in the agriculture and animal husbandry industry. It is mainly used in various important links of the whole animal husbandry industry chain, such as breeding, breeding, social services, slaughtering and processing, financial insurance, sales, operation and supervision. It is a service-oriented big data cloud computing platform that provides farmers and herdsmen, cooperatives, enterprises, etc. with feeding technology, breeding services, remote diagnosis, agricultural capital circulation, financial insurance, product sales, market analysis, operation and supervision services. (2024-05-16, Others, 0KB, 下载0次)
本文前三章分别简要介绍信息安全，入侵检测和安全日志的相关
概念和基本原理；第四章重点介绍数据挖掘技术，包括了关联规则分
析，序列分析，分类分析和聚类分析；第五章论证数据挖掘技术应用
于入侵检测系统的必要性和实施的系统架构；第六章详细说明实验流
程和实验过程中对关联规则算法的改进，其中重点研究了运用 K-均
值算法对审计日志做预处理和在实现关联规则挖掘时，如何改进
Apriori 算法，使之能在面对安全日志这种高维度数据时比原有算法
效率获得大幅度提高；
This article briefly describes the first three chapters were related to information security, intrusion detection, and security logs
The basic concepts and principles chapter focuses on data mining technology, including a sub-association rules
Analysis, sequence analysis, classification and clustering analysis Chapter demonstrate the application of data mining techniques
On the necessity of the intrusion detection system and system architecture implementation flow experiments described in detail in Chapter VI
Cheng and experiment to improve the algorithm of association rules, which are focused on the use of K-
Value preprocessing algorithm for audit logs and in the realization of association rule mining, the ways to improve
Apriori algorithm, so that it can log in the face of such high-dimensional data security than the original algorithm
Obtain greatly improved efficiency (2015-04-25, Others, 931KB, 下载1次)
：针对动力学方程未知且信噪比小的光混沌信号，采用小波多分辨分解算法对其进行噪音消
减．用Lorenz混沌信号对该算法的消噪效果进行了检验．提出利用互信息量法和Cao氏法来改进
小数据量法在时间延迟和嵌入维数计算上存在的主观选择性，对经过噪音消减的Lorenz混沌信号
利用此改进的小数据量法计算其最大Lyapunov指数．结果表明，信噪比可提高近10 dB左右，最
大Lyapunov指数计算误差可减少近30 ，并求得半导体放大器光混沌信号的最大Lyapunov指
数为0．389 6．
The wavelet multi—resolution decomposition algorithm was used for reducing noise of optical
chaos signals with dynamic equation unknown and low SNR．The algorithm was verified by Lorenz chaotic
signa1．The mutual information algorithm and Cao method were used to reduce the suhj ective in computing
the delay time and embedding dimension when applying the small data method to compute the largest
Lyapunov exponent．The largest lyapunov exponent of the de—noised chaos signal was calculated with this
improved method．The result shows that the SNR is increased by about 1 0 dB，and the error of the largest
Lyapunov exponent is reduced by 30 ．The largest Lyapunov exponent of the optical chaos signal 0．389 6
is obtained with this method． (2010-04-24, Others, 380KB, 下载18次)