没有合适的资源?快使用搜索试试~ 我知道了~
图与网络算法优秀论文171
试读
22页
需积分: 0 0 下载量 17 浏览量
更新于2022-08-03
收藏 3.83MB PDF 举报
在信息时代,如何理解信息的流动以及其对社会的影响已成为诸多学者关注的焦点。2016年数学建模竞赛(Mathematical Contest in Modeling,简称MCM)中,一篇题为“图与网络算法优秀论文171”的论文,以其创新的视角和深入的分析,为我们描绘了信息流动与信息价值之间的复杂关系,并提出了三种模型,以期解决信息传播中的若干问题。
在研究的起始阶段,为了更准确地捕捉信息流动的规律,作者构建了一个基于媒体影响力的SIIR模型。这个模型由经典的SIR模型发展而来,后者常用于模拟疾病的传播。本文将其应用于信息传播,将信息的传播过程类比于疾病的传播过程,以“热点传播节点”来模拟媒体在其中所起到的加速作用。通过对美国媒体数据的实证分析,发现该模型能够有效预测2050年信息网络的特征值,其预测性能令人瞩目。这一发现不仅验证了模型的适应性,也揭示了媒体在网络信息传播中的核心作用。
论文的第二部分,作者将研究的视角转向信息过滤问题,并将之视为一个分类问题。为了解决这一问题,作者设计了一种径向基函数(Radial Basis Function,RBF)网络。考虑到样本噪声对模型性能的影响,作者在实现RBF网络时采用了K-means聚类算法和最小均方误差(Recursive Least Squares,RLS)算法。通过在在线新闻数据集上进行实验,作者得到了相当高的训练准确度和测试准确度,分别为75.1%和74.3%。这表明对于新闻信息过滤任务,所提出的方法是行之有效的,且具有良好的实用前景。
在第三部分中,论文着重分析了社交网络中信息传播的特点,并建立了一个社交网络SIR(SN-SIR)模型。该模型的建立基于Facebook社交网络数据集,用以研究网络如何改变公众的兴趣和观点。通过模型模拟,作者分析了影响信息传播结果的多个因素,包括信息的价值、人们的初始观点、信息的形式或来源,以及网络的拓扑结构等。这为制定更有效的信息传播策略提供了理论依据。该模型有助于我们更深入地理解网络中信息传播的内在机制,以及这些机制是如何影响公众观念的形成和演变的。
论文的最后部分则关注了影响信息传播的外部因素,如政策、事件或其他外部刺激。这些因素能够在不同阶段加速或减缓信息在网络中的传播速度。通过对外部因素的深入分析,作者提供了一个更为全面的视角,使得信息传播的趋势预测更为准确,为信息的传播控制提供了策略上的参考。
这篇论文通过对信息流动的多维度研究,建立了三个创新的数学模型,并综合运用机器学习技术,对信息传播进行了全面的分析和预测。论文不仅在理论上为理解信息时代的数据传播特性提供了新的视角,而且在实践中为社交媒体和网络环境中的信息过滤与公众舆论的形成提供了科学的方法论。对于信息传播的理解,这篇论文无疑提供了一个富有深度和广度的研究范式,对于相关领域的研究与应用具有重大的启示作用。
For office use only
T1
T2
T3
T4
Team Control Number
44173
Problem Chosen
D
For office use only
F1
F2
F3
F4
2016 Mathematical Contest in Modeling (MCM) Summary Sheet
(Attach a copy of this page to each copy of your solution paper.)
How to Understand the Information
The flow of information has never been as easy or wide-ranging as it is today.
This paper proposes some models to explore the relationship between speed/flow
of information with inherent value of information.
First,to explore information flow, we build up a SIIR model which takes the me-
dia influence into consideration. This model is developed from the classical SIR
model.We introduce the concept of hot transmission nodes to highlight the impact
of media. We demonstrate our design with American media data. The results show
that our model has a significant performance in validating predicted values of the
characteristics of information network in 2050.
Second,the information filtering can be regarded as a classification problem. We
design a radial basis function network to implement the information filtering. In
order to solve sample noises and improve the computing performance,we decide
to use K-means algorithm and RLS algorithm to implement our RBF network. We
demonstrate our network in online news dataset ,with an training accuracy of
75.1%,and testing accuracy of 74.3%.we are satisfied that the accuracy rate with
information filtering as news.
Third, We have only considered the social network in the Internet and build a social
network SIR(SN-SIR) model to explore the problem-how public interest and opin-
ion can be changed through network.It simulates in the standard dataset: Facebook
social network dataset .By analyzing the factors which can implement our result-
s,we can make a plan to determinate information propagation. Then,we take infor-
mation values ,people ˛a´rs initial opinion and bias, form of the message or its source,
and topology or other reasons into account and propose a detailed scheme.
Finally, effects of external factors are considered in our models.And we analyze
the stability and sensitively of our models. Although there are some weakness in
our models,the results still demonstrate that our model can undergo disturbance in
certain extent.
数学中国
www.madio.net
本内容由645617861l同学与工作人员共同整理,恭喜同学成功完成数学中国第一期威客项目
数学中国:www.madio.net,最专业的数学建模平台
备战美赛,美赛资料下载:http://www.madio.net/forum-108-1.html
美赛讨论加入群:2017美赛官方群287336869
Team # 44173 Page 2 of 22
Contents
1 Introduction 3
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Our Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 Assumptions 4
3 A SIIR Model of Information Flow 4
3.1 Social Network Information Dissemination . . . . . . . . . . . . . . . . . . . . . . 4
3.1.1 Social Network Information Dissemination Characteristics . . . . . . . . . 4
3.1.2 Model Building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.2 Calculating and Simplifying the Model . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2.1 The Model Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2.2 Rationality validation And Sensitivity analysis . . . . . . . . . . . . . . . . 7
3.3 Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4 A Model of Information Filtering 11
4.1 Model Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2 Hybrid Learning Procedure For Networks . . . . . . . . . . . . . . . . . . . . . . . 11
4.3 Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.4 Sensitively Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5 A SN-SIR Model 14
5.1 The Way of Information Dissemination in Network . . . . . . . . . . . . . . . . . 14
5.2 A Model of Public Opinion Dissemination in Social Network . . . . . . . . . . . . 14
5.2.1 Model Building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.3 Sensitivity Analysis and Model Validation . . . . . . . . . . . . . . . . . . . . . . . 16
5.3.1 Model Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
5.3.2 Sensitivity, parameter validation analysis . . . . . . . . . . . . . . . . . . . 17
5.4 Result Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
6 Analysis 21
6.1 Strengths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6.2 Weaknesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
数学中国
www.madio.net
本内容由645617861l同学与工作人员共同整理,恭喜同学成功完成数学中国第一期威客项目
数学中国:www.madio.net,最专业的数学建模平台
备战美赛,美赛资料下载:http://www.madio.net/forum-108-1.html
美赛讨论加入群:2017美赛官方群287336869
Team # 44173 Page 3 of 22
1 Introduction
1.1 Background
Broadly speaking, information is one of the important ways for us to perceive the outside
world. Today it spreads quickly in tech-connected communication network. Sometimes it is
due to the information finding its way to influential or central network nodes that accelerate its
spread through social media. Our prevailing premise is that this cultural characteristic to share
information (both serious and trivial) has always been three. At present, information propa-
gation is exposed explosive growth.Beacause our social networks and medium are more and
more complex.Without effective ways to solve these problems ,it will finally cause information
propagation is more and more difficult to umderstand.Hence,our work has its unprecedented
significance in world nowadays.
1.2 Our Work
This paper propose some models to explore the relationship between speed/flow of informa-
tion with inherent value of information.
The Dilemma:
• An adaptive flow model is established with the consideration of the complex network
changes.
• Make use of the existing parameters to filter the information reasonably.
• Reasonable analysis of influencing factors in the process of information transmission.
The Approach:
• FirstˇcˇnTo explore information flow, we build up a SIIR model which takes the special fea-
tures if media influence into consideration. This model is developed from classical SIR
model.
• The Results show that our model has a significant performance in validate the value of
prediction with real value. And then ,we predict the rate of daily contact in 2050.
• Information filtering is regarded as classification problem. we design a radial basis func-
tion network to implement information filtering. In order to solve sample noise and com-
puting performance, we decide to use K-means algorithm and RLS algorithm to imple-
ment our RBF network.
• We have only considered the social network in the Internet and build a social network
SIR(SN-SIR) to explore the problem ˛a´s how public interest an opinion can be changed
through network ˛a´s.
• We can make a plan to determinate information propagation. Then ,we take information
value ,people ˛a´rs initial opinion and bias, form of the message or its source, and topology
or other reasons into account and build a detailed scheme.
数学中国
www.madio.net
本内容由645617861l同学与工作人员共同整理,恭喜同学成功完成数学中国第一期威客项目
数学中国:www.madio.net,最专业的数学建模平台
备战美赛,美赛资料下载:http://www.madio.net/forum-108-1.html
美赛讨论加入群:2017美赛官方群287336869
Team # 44173 Page 4 of 22
2 Assumptions
The accuracy of our models rely on certain key, simplifying assumptions. These assumptions
are listed below:
• Do not consider factors such as natural birth,death,and population mobility.The total
population keeps as a constant K.
• Infected individuals once contact with susceptible individuals,Infected individuals change
to susceptible individuals at a certain probability.
• At a certain time,The number of individuals per unit time removing from the infected
individuals is proportional to the number of patients.
• The number of media is relatively fixed in the same period.
Under the above and basic assumptions, we can set out to construct our model.
3 A SIIR Model of Information Flow
Information is created by the social network users and transmits in the entire network.The
result of transmission is that the receiver of the information translates into the transmitter and
disseminate the information to others,or translates to the recovered one and does not transmit
to others.Obviously, this transmission way is similar with the way Infectious diseases spread in
the crowd. Based on the analysis of social network structure,we introduce the hot transmission
node into network,and proposes an information transmission model based on the improved
SIR model [1],which is simulated with the data-based Facebook social network.
3.1 Social Network Information Dissemination
3.1.1 Social Network Information Dissemination Characteristics
In this section,we propose a to explore a general model of information flow problems, which is
not subjected to the limitations of the times and mediums of communication.
The total number of user nodes in social networks is denoted as N.The network is assumed to
be an undirected graph and each node in the network may be in one of three states: susceptible(s),
infected(i), or recovered(r).Nodes in state i have received the message and have the ability to
transmit it.Nodes in state s have not received the message and have the possibility to receive
it.Nodes in state r have received the message and will not transmit it any more.
There are a variety of mediums of communication,such as newspapers,televisions,the Inter-
net and so on.Different mediums have different ability of transmission,but all of these medi-
ums owe stronger ability than human beings.We consider media as the hot transmission n-
odes.Their contact rate is denoted as λ
1
. Ordinary users in networks are considered as ordinary
nodes.Their contact rate is denoted as λ
2
.
The laws of Information dissemination of social networks with hot transmission nodes are as
below:[2]
数学中国
www.madio.net
本内容由645617861l同学与工作人员共同整理,恭喜同学成功完成数学中国第一期威客项目
数学中国:www.madio.net,最专业的数学建模平台
备战美赛,美赛资料下载:http://www.madio.net/forum-108-1.html
美赛讨论加入群:2017美赛官方群287336869
剩余21页未读,继续阅读
资源推荐
资源评论
198 浏览量
106 浏览量
2022-08-03 上传
2022-08-03 上传
193 浏览量
179 浏览量
2022-08-03 上传
2022-08-04 上传
173 浏览量
166 浏览量
179 浏览量
172 浏览量
196 浏览量
166 浏览量
146 浏览量
120 浏览量
2020-02-19 上传
5星 · 资源好评率100%
187 浏览量
2010-06-07 上传
2020-02-19 上传
158 浏览量
172 浏览量
102 浏览量
119 浏览量
136 浏览量
资源评论
MurcielagoS
- 粉丝: 20
- 资源: 319
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
- 条件方差总和 (SCV) 是多模态图像配准的新相似性指标Matlab代码.rar
- 通过 FFT 按总(广义)变化进行图像降噪Matlab代码.rar
- 填充圆圈或麦田圈区域Matlab代码.rar
- 通过 RGB-HSI 图像融合提高图像空间分辨率的 GUI Matlab.rar
- 通过 HSV 色彩空间中的阈值检测图像中的彩色区域。.rar
- 通过 RGB-HSI 图像融合提高图像空间分辨率的 GUI MATLAB代码.rar
- 通过使用强度平均值度量从用户定义的种子点生长区域来进行分割。.rar
- 通过使用强度平均度量从种子点生长区域进行分割Matlab代码.rar
- 通过基于 Adaptive Cuckoo 搜索的 Wiener 滤波器进行多光谱图像去噪Matlab代码.rar
- 通过在矩形波导中设计耦合腔体滤波器Matlab代码.rar
- 图像 2D 变换 平移 旋转 缩放Matlab代码.rar
- 图像的 Choatic Baker 映射解密Matlab代码.rar
- 图像的对比度调整Matlab代码.rar
- 通过自适应核回归实现视频超分辨率Matlab代码.rar
- 图像的哈希值Matlab代码.rar
- 图像对比度拉伸Matlab代码.rar
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功