分类问题的在线顺序双并联快速学习机

双并联前馈神经网络模型是单层感知机和单隐层前馈神经网络的混合结构,本文构造了一种双并联快速学习机算法,与其他类似算法比较,提出的算法能利用较少的隐层单元及更少的待定参数,获得近似的学习性能.数值实验表明,对很多实际分类问题,提出的算法具备更佳的泛化能力,因而可以作为快速学习机算法的有益补充.     

货车枕簧丢失故障动态图像识别方法

针对货车运行故障动态图像检测,提出无故障目标识别工作模式,解决货车枕簧丢失故障的自动识别问题。利用Haar特征提取枕簧特征信息,基于AdaBoost算法选取特征并构建层叠分类器,等比缩放搜索窗口检测货车图像,最终分选出无故障的枕簧图像,从而大大地减少了待识别图像的数量,显著地提高了人工识别效率。实验表明,该算法使用的特征简单,搜索策略高效,不受枕簧位置、缩放和旋转的影响,抗噪能力强,对分辨率低、局部遮挡、光照不足或过度曝光等质量较差的图像仍具有很强的适应性,所提出的方案能够满足全天候条件下的货车枕簧目标识别,为货车故障动态图像检测的工程化应用奠定了基础。     

基于6S模型的遥感影像逐像元大气纠正算法

大气纠正的目的是从遥感影像中去除大气影响,并反演获取地物真实反射率。介绍了一种逐像元对遥感影像进行大气纠正的算法,该算法基于6S(Second Simulation of the Satellite Signal in the Solar Spectrum)大气辐射传输模型计算建立的查找表(look-up table),并利用地面暗目标(dark object)进行陆地气溶胶光学厚度的自动反演,由于气溶胶的分布具有空间连续性,在获取地面暗目标气溶胶光学厚度值后,通过空间插值的方法计算影像中非暗目标像元的气溶胶光学厚度值,经过查找表二次插值计算,逐像元进行大气纠正并获取像元地表反射率值。以Landsat5遥感影像为例,介绍了算法流程,展示了大气纠正的结果。结果显示,利用查找表逐像元大气纠正的算法,能够在一定程度上去除云雾对影像的影响,更加精确的对遥感影像进行大气纠正并获取地物的真实反射率。     

低曝光量下投影算法的稳健性研究

研究了在低曝光量情况下灰度投影算法的稳健性,为提出的稳定成像系统的设计与实现提供可行性论证。通过合成视频实验探讨、实际视频实验论证相结合的研究方法,探讨曝光量不足对灰度投影算法性能的影响。实验结果表明,灰度投影算法在欠曝5档和存在一定随机噪声的情况下,仍能保持良好的运动矢量估计性能,从而说明了投影算法在一定的曝光量范围内仍能准确地得到运动矢量估计,是一种运算量小、实时性好且稳健性较强的电子稳像算法。投影算法在低曝光量情况下良好的稳像稳健特性,使其能在低亮度及高实时性要求等环境下应用于航拍、卫星遥感等稳定成像或高清晰成像场合。     

现代数字物理教学连载④——工科基础量子教学中如何发挥数字化技术的作用

介绍了如何运用数字技术的三维表现手法和能力以及数值计算与模拟功能来深化量子理论的基本概念与基础内容的教学,充实了量子理论实际应用的教学内容.     

现代数字物理教学连载⑩——在数字技术平台上的研究式学习

介绍了大学物理教学中,如何利用数字技术平台引导学生进行自主的研究式学习.首先是知识的拓展与转移,接着是教师指导下的研究与探索,最后是学生自主独立的研究与创新.实践证明,我们创造的这一教学方法对于培养学生的科学素质与能力十分奏效.     

《光学学报》四十周年:温故开新,不惑笃行

自1981年创刊以来,《光学学报》以发表国内优秀光学科研成果为己任,为我国光学科技人员提供了高水平、活跃的学术交流平台.《光学学报》经历了近40年的洗礼,见证了国内一代代优秀光学科研人员的成长.本文首先回顾了《光学学报》从创刊到转企、集群化发展,再到数字化出版的变革之路;其次,统计了《光学学报》历年来的出版数据、数据库...     

An Edge Server Placement Method Based on Reinforcement Learning

In mobile edge computing systems, the edge server placement problem is mainly tackled as a multi-objective optimization problem and solved with mixed integer programming, heuristic or meta-heuristic algorithms, etc. These methods, however, have profound defect implications such as poor scalability, local optimal solutions, and parameter tuning difficulties. To overcome these defects, we propose a novel edge server placement algorithm based on deep q-network and reinforcement learning, dubbed DQN-ESPA, which can achieve optimal placements without relying on previous placement experience. In DQN-ESPA, the edge server placement problem is modeled as a Markov decision process, which is formalized with the state space, action space and reward function, and it is subsequently solved using a reinforcement learning algorithm. Experimental results using real datasets from Shanghai Telecom show that DQN-ESPA outperforms state-of-the-art algorithms such as simulated annealing placement algorithm (SAPA), Top-K placement algorithm (TKPA), K-Means placement algorithm (KMPA), and random placement algorithm (RPA). In particular, with a comprehensive consideration of access delay and workload balance, DQN-ESPA achieves up to 13.40% and 15.54% better placement performance for 100 and 300 edge servers respectively.     

A Banking Platform to Leverage Data Driven Marketing with Machine Learning

Payment data is one of the most valuable assets that retail banks can leverage as the major competitive advantage with respect to new entrants such as Fintech companies or giant internet companies. In marketing, the value behind data relates to the power of encoding customer preferences: the better you know your customer, the better your marketing strategy. In this paper, we present a B2B2C lead generation application based on payment transaction data within the online banking system. In this approach, the bank is an intermediary between its private customers and merchants. The bank uses its competence in Machine Learning driven marketing to build a lead generation application that helps merchants run data driven campaigns through the banking channels to reach retail customers. The bank’s retail customers trade the utility hidden in its payment transaction data for special offers and discounts offered by merchants. During the entire process banks protects the privacy of the retail customer.     

A Hybrid Method Using HAVOK Analysis and Machine Learning for Predicting Chaotic Time Series

The prediction of chaotic time series systems has remained a challenging problem in recent decades. A hybrid method using Hankel Alternative View Of Koopman (HAVOK) analysis and machine learning (HAVOK-ML) is developed to predict chaotic time series. HAVOK-ML simulates the time series by reconstructing a closed linear model so as to achieve the purpose of prediction. It decomposes chaotic dynamics into intermittently forced linear systems by HAVOK analysis and estimates the external intermittently forcing term using machine learning. The prediction performance evaluations confirm that the proposed method has superior forecasting skills compared with existing prediction methods.