Time Averaging Algorithms with Stopping Rules for Multi‐Agent Consensus with Noisy Measurements

A time averaging technique is introduced to consensus algorithms in networked multi‐agent systems under a noisy environment. Each agent communicates with its neighboring agents via a constant gain, while the time averaging states of the agents are considered as the values for agreement. The variation of the time averaging states then is evaluated at a specific number of iterations, where the number is given explicitly in terms of parameters related to the consensus accuracy and its probabilistic guarantee. This result establishes a rigorous stopping rule for the multi‐agent consensus with noisy measurements. Several results on this type of stopping rules are provided for undirected, directed, and time‐varying communication graphs. These theoretical results are illustrated through numerical examples.     

Mining fuzzy temporal association rules by item lifespans

Data mining is the process of extracting desirable knowledge or interesting patterns from existing databases for specific purposes. In real-world applications, transactions may contain quantitative values and each item may have a lifespan from a temporal database. In this paper, we thus propose a data mining algorithm for deriving fuzzy temporal association rules. It first transforms each quantitative value into a fuzzy set using the given membership functions. Meanwhile, item lifespans are collected and recorded in a temporal information table through a transformation process. The algorithm then calculates the scalar cardinality of each linguistic term of each item. A mining process based on fuzzy counts and item lifespans is then performed to find fuzzy temporal association rules. Experiments are finally performed on two simulation datasets and the foodmart dataset to show the effectiveness and the efficiency of the proposed approach.     

大数据关联关系度量研究综述

大数据关联性分析是大数据挖掘的基础,一个好的关联性度量是实施关联分析的关键。本文首先指出大数据时代关联度量面临的挑战和研究现状,从关联关系度量的构造角度出发,对现有的关联关系度量进行整理,归纳总结了这些关联关系的性质和适用条件。在回顾关联度量发展历程的基础上,结合大数据时代关联关系的特点,提出构造关联度量可能满足的条件。最后针对多模态数据关联关系度量的若干问题进行探讨和梳理,从3个角度出发,提出应对多模态数据空间转换的挑战,以引起对该领域更深入的思考与研究工作,从而促进大数据挖掘工作的进展。     

Protel DXP 2004应用中的常见问题及解决方法

Protel DXP 2004软件是一款应用广泛的印制电路板设计软件,学习该软件比较容易入门,但是要熟练掌握并且灵活处理软件使用中的常见问题并非易事。从电气规则检查、载入网络表、制作元件封装等方面,分析Protel DXP 2004软件操作中的常见问题及产生原因,并提出了解决这些问题的方法。     

Style synthesis and analysis of car designs for style quantification based on product appearance similarities

Understanding how similar design appears is a key element to understanding companies’ design strategies. However, it is difficult to evaluate companies’ design strategies with conventional style measurement methods since they only taxonomically measure whether a specific characteristic is included in a specific style. This study numerically measured car design similarities to synthesize and analyze car brand styles, thereupon discovering the design trends among car brands for strategic design positioning. This paper aims to find methods for quantifying style differences and identifying unique design elements of car designs among 23 automobile manufacturers based on design similarities of a large quantity of car designs (N = 119). To achieve this goal, a hybrid style quantification methodology – a mixture of Fourier decomposition, eye tracker, and shape grammar – was created to evaluate similarities, visual significance, and combinations of 19 car design elements. Fourier decomposition was incorporated to find the quantifiable values of design similarities of car design elements. Visual significance analysis was also conducted for each car design element through eye tracker to measure the importance of certain design elements for weighting factors. Then, each combination of design elements was compared with car design elements of other cars for similarity calculations. Finally, car design alternatives were synthesized, and transitions of design positioning were analyzed based on the similarity values weighed by the visual significance results. Using the suggested methods, alternate designs can be synthesized while preserving brands’ design styles, and design trends can be analyzed for strategic evaluation.     

数据挖掘技术在教育学中的应用

随着信息化时代的不断发展,数据挖掘技术日趋成熟,满足了人们对于大量信息的处理要求。目前,数据挖掘技术已经越来越多的应用于金融、通讯和交通等各行各业,但是在教育领域的应用相对较少。针对这一现状,在传统的分析方法上采用了关联规则挖掘,聚类挖掘等多种算法,对青少年同伴关系、人际关系和网络成瘾等各方面进行研究,得到不同性别的青少年同伴亲疏程度不同等一系列结论,提出了数据挖掘技术在教育学领域的应用的新前景。     

校园网安全防护的研究

目的:探讨校园网组建起来后,所面临的各种问题,并提供有效的解决方案。方法:从网络的结构性,网络的可靠性,网络的安全性三个方面进行分析,并提出相应的优化策略。结果:提高了校园网运行的可靠性、安全性和效率。结论:采用路径分析、关联规则待方法对网络进行优化,可以提高校园网的运行效率以及安全性。     

Pedestrian Detection and Tracking Using Deformable Part Models and Kalman Filtering

Abstract： Pedestrian detection techniques are important and challenging especially for complex real world scenes. They can be used for ensuring pedestrian safety, ADASs （advance driver assistance systems） and safety surveillance systems. In this paper, we propose a novel approach for multi-person tracking-by-detection using deformable part models in Kalman filtering framework. The Kalman filter is used to keep track of each person and a unique label is assigned to each tracked individual. Based on this approach, people can enter and leave the scene randomly. We test and demonstrate our results on Caltech Pedestrian benchmark, which is two orders of magnitude larger than any other existing datasets and consists of pedestrians varying widely in appearance, pose and scale. Complex situations such as people occluded by each other are handled gracefully and individual persons can be tracked correctly after a group of people split. Experiments confirm the real-time performance and robustness of our system, working in complex scenes. Our tracking model gives a tracking accuracy of 72.8% and a tracking precision of 82.3%. We can further reduce false positives by 2.8%, using Kalman filtering.     

调亏灌溉条件下甘蔗耗水规律试验研究

通过大棚内的甘蔗盆栽试验,研究了不同生育期调亏灌溉条件下甘蔗的耗水规律。结果表明:甘蔗的耗水强度因土壤含水量的不同而不同,各生育期内,土壤含水量大的甘蔗耗水强度就大;甘蔗各生育期的阶段耗水量因生育期的长短而不同,萌芽期的甘蔗耗水量最小,伸长期的耗水量最大。在不考虑降雨因素影响的条件下,研究了甘蔗生长的耗水规律参数。     

一种利用关联规则的改进朴素贝叶斯分类算法

朴素贝叶斯分类是一种简单而高效的分类模型,然而条件独立性假设在现实中很少出现,致使其性能有所下降。通过引入关联规则,从两方面来改善朴素贝叶斯分类的性能。一方面,通过对关联规则的挖掘,发现条件属性之间的关联关系,并且利用这种关联关系弱化朴素贝叶斯的独立性假设;另一方面,通过关联规则的置信度,给朴素贝叶斯加权。