Пікірлер
@FatBitches
@FatBitches 9 ай бұрын
thank you
@IreneTafani
@IreneTafani Жыл бұрын
i love this guy
@IreneTafani
@IreneTafani Жыл бұрын
How can we find the correct values for the different parameters? c, C and delta?
@IreneTafani
@IreneTafani Жыл бұрын
Thank you so much!
@andrewd4413
@andrewd4413 Жыл бұрын
Being surrounded by people who look like you and who are more willing to associate with you because you share their race (based on appearance) is a huge advantage that white Americans do not ever realize unless they grew up in a place surrounded by people who were not white. When I went to my mother's hometown in central Mexico, I saw so many people who looked like me, and it was so shocking and strange to experience it. My entire life, I have been surrounded by communities and schools that are over 80% White, and even then, there are very few people who are Mestizo and/or Latin American like me. People like to pretend to be colorblind and dismiss race, but racial homophily is subconscious and does exist even among young children across all races. It's a fact that people tend to associate with people who look like them, and as an only child of a Mestiza Mexican mother and White Anglo-American father, it has been very isolating. So many times in my life, I have tried to develop close bonds with white friends, but they did not work out as well as friendships I made with fellow people of color. I somehow always knew subconsciously that white people were less willing to accept me because of my race, and it is only until recently that I consciously realize it. It is subconscious for them. White Americans in most of Missouri have the privilege of never having to interact with people of a different race from them if they choose too but if I want to get ahead in life I have to interact with them even if they never fully accept me as one of them. They subconsciously don't want non-Whites in their social group no matter how hard we try to assimilate and act like them. I will not abandon my Mexican heritage and Mestizo identity to try to get White Americans to accept me because no matter what I do, they never will. I'm tired of the extreme lack of self-awareness and introspection of white Americans to put themselves in the shoes of a racial minority in an overwhelmingly white school that also happens to be mixed-race. Do they ever think for even a second about what it must be like to grow up being the only one or one of the few of your race surrounded by more privileged people of another race? It matters because people consciously and subconsciously treat you differently based on your race and appearance. It makes networking and friendships that much harder. Do they ever check their subconscious bias of racial homophily and consider trying to trach out and make friends with racial minorities?
@Rohleksh
@Rohleksh Жыл бұрын
Much needed video!
@adeparanata7270
@adeparanata7270 Жыл бұрын
Amazing... really clear. Big thanks!
@muskduh
@muskduh Жыл бұрын
Thanks for the video.
@NeotenicApe
@NeotenicApe 2 жыл бұрын
I love this man so much, thank you ❤️
@muskduh
@muskduh 2 жыл бұрын
Thanks for the video!
@muskduh
@muskduh 2 жыл бұрын
Thanks for the video
@페퍼로니는더블
@페퍼로니는더블 2 жыл бұрын
Hello, professor. I would like to say I really appreciate your upload. I have a small question while I am listening to this: is there any different between big N and small n? I know that N is the number of nodes, but I am not sure about small n. Is n used indifferently from N? or does it have any different meaning?
@muskduh
@muskduh 2 жыл бұрын
Thanks very much for your time
@muskduh
@muskduh 2 жыл бұрын
This is so good thanks a million
@muskduh
@muskduh 2 жыл бұрын
appreciate the video
@mohamedfouad6492
@mohamedfouad6492 2 жыл бұрын
stanford course?
@zhangsc91
@zhangsc91 2 жыл бұрын
Thank you so much for your mini series on ERGM!
@BroadcaststoNowhere
@BroadcaststoNowhere 3 жыл бұрын
Thank you for making this information available.
@BroadcaststoNowhere
@BroadcaststoNowhere 3 жыл бұрын
kzbin.info/www/bejne/jXy4lnWlhtVli9E
@SillyDuckyLuck
@SillyDuckyLuck 3 жыл бұрын
you are a king!
@SillyDuckyLuck
@SillyDuckyLuck 3 жыл бұрын
top!
@mullactalk
@mullactalk 3 жыл бұрын
Very happy to find this video :) I'm doing a project on DeGroot and this is a great introduction
@ВасилСтойчев-к8х
@ВасилСтойчев-к8х 3 жыл бұрын
Thank you! Very useful!
@nickfrancis7292
@nickfrancis7292 3 жыл бұрын
Brilliant, clear and powerful lecture.
@flornotlikeyou
@flornotlikeyou 4 жыл бұрын
i love the way he explains everything, he makes it pretty easy to understand :), really interesting channel
@mariamjf7677
@mariamjf7677 4 жыл бұрын
Thank you for this!
@vanisher49
@vanisher49 4 жыл бұрын
Thank you this helped me a lot!
@Iceport
@Iceport 4 жыл бұрын
This really helps build up my intuition. Thanks!
@snehalkanade6266
@snehalkanade6266 4 жыл бұрын
Sir, this is a very good course and I was just thinking that can we have a test at the end of this 7 weeks course provided with certification
@jackson.matthewo
@jackson.matthewo 4 жыл бұрын
You can find a version of the course on coursera that has problem sets, a final exam, a certificate, and some additional empirical materials: www.coursera.org/learn/social-economic-networks/home/welcome
@zinanliu9212
@zinanliu9212 4 жыл бұрын
Can you explain why decay centrality when delta is 0 is direct connected to all the other node and delta = 1 is indirect connected to the rest?
@fionayang1112
@fionayang1112 4 жыл бұрын
thx professor! 注:非专业翻译,本人菜鸡。 p1 回顾一下,ergm模型由权重乘以统计量的和表示。这一形式使得我们能知道图里有什么特征,哪些特征是显著的,图形成的主要原因是什么。 p2 florentine marriages(反之类似一个家族商业联姻网络) 这篇文章的想法:统计了一下florentine marriages里各种统计指标(如link,2-星,3-星,三角形),然后estimate图的概率表达式中有哪些系数(就是那些权重beta)和随机图比是显著的,然后就能知道是哪些指标造成了网络的形成? p3 estimate的算法逻辑:已知观测网络,在最大化该网络出现的概率时,这些系数是什么?(就是一个最大似然估计) p4 p5 florentine families的结果(应该是做完scaling的结果) 可以看到link的系数是-4.27,而标准差是1.13,也就是说在3个标准差外,并是负数,这说明,在这个网络中的link比起随机图来说是显著的少的,link并不是这个网络的主要特征。 类推,triad的系数是1.32,而标准差是0.65,也就是说在2个标准差外,并是正数,这说明,在这个网络中的triad比起随机图来说是显著的多的,triad是这个网络的主要特征。 p6 ergm模型测量的难点在于计算标准化需要的那个分母,也就是所有可能出现的图的指标之和难以计算,因为图太多了,设想一个30个节点的图,就有2^(30*29/2)种图,这么多的图根本无法穷尽。那么抽样呢?所以科学家采用了MCMC方法在图上随机走动,并期待多次随机走动的得到的样本结果是具有代表性的,可以代表总体的图的情况。 p7 然而,mcmc的方法并不是有效的,一项研究指出,对于一个稠密的网络,只有当网络中有相当多的独立link时,结果才会在想对较短的时间内收敛到精确的数字。这就意味着,似乎是从node-node的层次去理解这些网络,而不是从网络自身去理解,而从node-node层次去理解网络已有其他的技术,不能体现出ergm的长处,如果ergm不能从网络层面的特征理解网络,不能做到和其他现有技术不一样,那ergm有什么用呢? p8 例子:用mcmc方法,估计ergm模型中的系数,可能会出现的问题: 模拟一下网络,30个节点,10个独立,另外20个节点中,有65条link。这个网络出现的概率用ergm模型表示,beta1*link数 + beta2*isolate孤立点数。真实的beta1是-0.65,真实的beta2是-0.69。接下来用mcmc模拟估计两个beta的数值。模拟25次,每次中都有10000次mcmc run(样本数量10000)。结果显示: p9,p10 对link的beta1的估计一般,岁isolate点的beta2估计完全无效。 p11 mcmc技术的estimate是不准确的。
@fionayang1112
@fionayang1112 4 жыл бұрын
Thx a lot professor Matthew !! I did a very simple translation in Chinese, i hope you do not mind it. I am not professional but hope it helpful: p1 随机指数图模型的优点在于: 1. 包括更多更丰富的网络局部特征(local features)和相关性 (dependencies) - 不仅是cluster, 三元组,而是细小到clique,isolate node。 2. 方便统计估计 - 随机指数家族有良好的统计性质 p2 从历史的角度看:ergm曾被称为马尔可夫模型,p星模型。 (插播:哈哈哈哈intersting, 因为之前的模型都不够好,所以科学家搞了一堆hybrid model混血模型) 我们想要的是:link ij 形成的概率依赖于link jk 和 link lk(三元闭合)。但是这样面临了一个问题:一旦我们让边的形成相互依赖,那么以为着一切边的形成都是关联的correlated,边变成互锁的interlock,一条边影响它的邻居,它的邻居又反过来影响它(类似内生性问题??),于是我们面对的这个系统将是一个拥有大量相关性的系统。我们需要去识别整体的关联性。 p3 一个简单的例子: 1. 网络的概率 等于 网络中形成的边数量 2. 网络的概率 等于 网络中形成的三元组数量 (试想:拥有相同的边的数量,而不通过三元组数量的两个网络) p4 边的似然,取决于节点的属性,也取决于节点是否有共同的邻居。 p5 网络G的概率 依赖于 系数1*G拥有的边的数量 + 系数2*G拥有的三元组的数量 因为上述式子可能会出现负数,而概率不可能是负数,为了式子有更好的性质,所以加上exp,让图的概率可以一直是正的。 p6 在 hammer sly 1971未发表的论文中,他指出:任何的网络模型,都可以表示为指数家族with 图中的统计上的加总。 (试想,假设你有一个很复杂的具体模型,你可以写上很多很多的统计量:边,三元组,二元组,clique,点的属性,出入度分布等等等等) p7 给定网络G的概率 等于 exp(系数1*G拥有的边的数量 + 系数2*G拥有的三元组的数量) 除以(我理解下是)所有的和G有相同的节点和属性的可能的网络的 exp(系数1*G拥有的边的数量 + 系数2*G拥有的三元组的数量) 的和。(I am not pretty sure...) p8 endos-renyi gnp 为例子,最后的s1(g)是statistic , L(g), g拥有的边的数量
@samsang8971
@samsang8971 4 жыл бұрын
Thank you Professor! Your videos helped me a lot!
@dextergavin6950
@dextergavin6950 3 жыл бұрын
sorry to be so offtopic but does any of you know of a tool to log back into an instagram account?? I somehow forgot my login password. I would appreciate any help you can give me.
@pedrotucker191
@pedrotucker191 3 жыл бұрын
@Dexter Gavin instablaster =)
@dextergavin6950
@dextergavin6950 3 жыл бұрын
@Pedro Tucker Thanks for your reply. I found the site through google and im in the hacking process now. Seems to take quite some time so I will get back to you later with my results.
@dextergavin6950
@dextergavin6950 3 жыл бұрын
@Pedro Tucker it worked and I now got access to my account again. I'm so happy:D Thank you so much you saved my ass!
@pedrotucker191
@pedrotucker191 3 жыл бұрын
@Dexter Gavin Glad I could help :D
@DJBaBoresds
@DJBaBoresds 5 жыл бұрын
Great information
@nourabsi5836
@nourabsi5836 5 жыл бұрын
I have no idea why someone would give you a thumbs down
@markkreikamp6295
@markkreikamp6295 5 жыл бұрын
Is it possible to add interaction terms to a valued ERGM?
@markkreikamp6295
@markkreikamp6295 5 жыл бұрын
is it possible to add interaction terms with either structural variables or other variables to valued ERGMS?
@fanyin9056
@fanyin9056 5 жыл бұрын
Most of the real world graphs are sparse
@tobiashellinger2165
@tobiashellinger2165 5 жыл бұрын
how to calculate eigenvector centrality =??=
@socialandeconomicnetworks4586
@socialandeconomicnetworks4586 5 жыл бұрын
It is a solution to a simultaneous set of equations and something that many software programs (both network software and others) can provide for moderate size networks, or approximate for larger networks. There are many eigenvectors for such a matrix, but generally a unique one (up to a rescaling) with all non-negative values for networks that we deal with in practice.
@johnsundquist6960
@johnsundquist6960 6 жыл бұрын
At 14:10 this should be .528 (not .628), I'm pretty sure.
@codexhammered007
@codexhammered007 6 жыл бұрын
bhai kyu chutiya bana rha baithe baithe? :-/
@ng2530
@ng2530 4 жыл бұрын
rehne de teri aukaat nahi hai
@_chloekwak1803
@_chloekwak1803 6 жыл бұрын
Thank you so much for your lessons, Professor! It was a long, great journey with you.
@_chloekwak1803
@_chloekwak1803 6 жыл бұрын
Nice studio tour! Thank you Professor! :)
@bens4446
@bens4446 6 жыл бұрын
Please provide a reference to a paper explaining the "2nd eigenvalue of the stochastized matrix" thing. Or anything on the meaning of adjacency matrix eigenvalues in general.
@socialandeconomicnetworks4586
@socialandeconomicnetworks4586 6 жыл бұрын
Hi You can find more background materials and references for details not in the videos on the Coursera course site: www.coursera.org/learn/social-economic-networks/home/welcome and also on my website: web.stanford.edu/~jacksonm/ best, Matt
@DrZenphy
@DrZenphy 7 жыл бұрын
How to calculate the eigenvector centrality?
@estevaoalves5614
@estevaoalves5614 6 жыл бұрын
I would like to know it too
@rohitshetgar6929
@rohitshetgar6929 7 жыл бұрын
Can you make a video on chromatic polynomials ,,, covering the important topics of it and few theorems related to it .
@BroadcaststoNowhere
@BroadcaststoNowhere 3 жыл бұрын
kzbin.info/www/bejne/gGjKZnuDj8x-g7c
@BroadcaststoNowhere
@BroadcaststoNowhere 3 жыл бұрын
kzbin.info/www/bejne/kGqzdaSJZ9x_qKs
@Rohit-nb8nf
@Rohit-nb8nf 3 жыл бұрын
@@BroadcaststoNowhere Thank you for the reply after 3 years , I finished my Degree itself 😀
@Chris-wy4rr
@Chris-wy4rr 7 жыл бұрын
very nice explanation
@namdoanduc
@namdoanduc 7 жыл бұрын
Greate lecture! Please public this slide.
@lilianacastro7934
@lilianacastro7934 7 жыл бұрын
Very clear!!! Thank you,
@garrettfernando7415
@garrettfernando7415 3 жыл бұрын
I guess Im randomly asking but does any of you know of a trick to get back into an Instagram account..? I stupidly forgot the password. I appreciate any help you can offer me
@lyleelias8595
@lyleelias8595 3 жыл бұрын
@Garrett Fernando instablaster :)
@garrettfernando7415
@garrettfernando7415 3 жыл бұрын
@Lyle Elias i really appreciate your reply. I got to the site through google and im waiting for the hacking stuff atm. I see it takes quite some time so I will get back to you later when my account password hopefully is recovered.
@ANUBHAVSHRESTHABCE
@ANUBHAVSHRESTHABCE 7 жыл бұрын
Tatti
@ChiranjibPatra
@ChiranjibPatra 7 жыл бұрын
Nice explanation