N*********6
发帖数: 2302 | 1 【 以下文字转载自 Parenting 讨论区 】
发信人: lilycola (妞妞猪), 信区: Parenting
标 题: [合集] Re: 我对ＡＡ的看法
发信站: BBS 未名空间站 (Wed Mar 25 17:34:03 2015, 美东)
☆─────────────────────────────────────☆
northbeach ((想认识一个小时候放过羊的朋友)) 于 (Tue Mar 24 16:27:55 2015, 美东) 提到:
“在美国, 所谓“affirmative action” 又叫“平权法案”，指的是“平等人权在就业
领域的体现”，即 “equal opportunity employment”。具体而言，“平权法案”要
求美国联邦的各级 contractors and subcontractors必须遵照执行。 其目的是杜绝各
类就业歧视，包括因肤色，信仰，性别，国别等引起的歧视。”
上面是关于ＡＡ的解释，来源来自旁边的楼。从字面上，ＡＡ绝对是个好东西，让各个
种族都能被平等地对待。 同时，ＡＡ在美国历史上正面作用不可抹杀。... 阅读全帖 |
|
v***r
发帖数: 1046 | 2 correlation存在，causality需要去研究。版上讨论的没有人一定拿了correlation做
证据仓促下结论。你知道个correlation and causality区别，就车轱轳套话一遍又一
遍。没有不硬期吗？晚饭吃的人参配罗卜吗? |
|
v***r
发帖数: 1046 | 3 不胜惶恐。颂读几个关键词，希望在您老心目中加点分:
"文科生，截面, correlation, two-particle system"
"文科生，截面, correlation, two-particle system"
"文科生，截面, correlation, two-particle system" |
|
c***z
发帖数: 6348 | 4 【 以下文字转载自 Military 讨论区 】
发信人: chaoz (没钱也任性), 信区: Military
标 题: Re: 机器学习是怎么回事
发信站: BBS 未名空间站 (Wed Jan 6 19:33:46 2016, 美东)
Do you know calculus?
Say you have a function y = f(x), where f has some parameters, and follow
some basic requirement (e.g. continuous, but this is not necessary, because
saw shaped functions work better).
Basically it is using the chain rule in calculus to move the parameters
around for minimal prediction error, which is the difference between the
predicted y and real y.
... 阅读全帖 |
|
d**********i
发帖数: 524 | 5 【 以下文字转载自 Military 讨论区 】
发信人: dayuguanshui (大禹灌水), 信区: Military
标 题: 基因认为我是自由派 不是保守派
发信站: BBS 未名空间站 (Sat Feb 3 09:53:44 2018, 美东)
We did find a significant association between political conservatism and
rs10952668 (Table 5). This SNP lies in LOC642355, a pseudogene on chromosome
7. Not surprisingly, the SNP also showed an association with the highly
correlated trait of Democrat versus Republican (b=0.260, P<
.02).
An interesting finding was that the SNP associated with political
cons... 阅读全帖 |
|
e*******e
发帖数: 6165 | 6 这么说吧，网上的性格和现实肯定是有差距的，但至少是correlated，是strong
correlated 还是 weak correlated 就不一定了。
因此是不是可以从这人想要表现的形象来反推她大致真实的样子？ |
|
l****g
发帖数: 5080 | 7 The example is not good. For diamond, the quality is not correlated with
price, the quality is correlated with the perception mostly. For the 4C
measurements, lab created one is better than earth created, yet cheaper.
Ultimately quality is correlated with the perception. But for most stuff,
more expensive one means better quality, but not for diamond, aluminum oxide
(often called sapphire), arts, etc.
|
|
c****t
发帖数: 19049 | 8 Some particles such as photons and electrons, when brought together and then
separated, can become “entangled” and exist in the same quantum
mechanical state.
This quantum mechanical state, prior to any form of measurement, possesses
indefinite properties such as position in space, spin, and polarization.
Such properties remain indefinite until they are measured, at which point
they take on definite values such as a particular location in space or a
clockwise spin. After one such entangled parti... 阅读全帖 |
|
B******1
发帖数: 9094 | 9 How 相干??
I claim that "the year 2012 is a dragon year" is as 相干 as the
investigation to the improvement of the admission rate! This just shows
the absurdity of putting those two statements (events) together!
Logical fallacy: correlation vs. causation
Post hoc ergo propter hoc, Latin for "after this, therefore because of this"
, is a logical fallacy (of the questionable cause variety) that states "
Since that event followed this one, that event must have been caused by this
one." It is often shor... 阅读全帖 |
|
s******8
发帖数: 2131 | 10 for example, LibSVM calculates the weights of each attribute. How to
intepret those weights? some are positive numbers and some are negative.
Does it mean 'correlation" to the class label? like positively correlated
and negatively correlated?
thx. |
|
S**Y
发帖数: 136 | 11 不大明白
能详细说下吗?
马尔科夫模型state是什么? 高斯模型用来模拟什么?
每只鸭子的天与天之间的correlation不同,你怎么量化correlation强的鸭子,怎么量化
correlation弱的鸭子?
谢谢. |
|
s******y
发帖数: 28562 | 12 我觉得他的文章的意思就是大部分基因的表达程度都和细胞的活跃程度有关，
所以随便乱抓几个都能看出区别来，但是和癌症的机理基本上就是打酱油的关系。
总体意思就是目前的主流方法找出来的marker不能作为推测机理的起点。
他们在faculty 1000上抱怨说他们这篇文章一共投了4年，前后被六个
顶级杂志据了（他们直接点名的几家估计是对他们态度最糟糕的，哈哈哈）
这里是他们在PLoS Comp Biol 发表的文章的摘要。
Bridging the gap between animal or in vitro models and human disease is
essential in medical research. Researchers often suggest that a biological
mechanism is relevant to human cancer from the statistical association of a
gene expression marker (a signature) of this mechanism, that was d... 阅读全帖 |
|
s******y
发帖数: 28562 | 13 我觉得他的文章的意思就是大部分基因的表达程度都和细胞的活跃程度有关，
所以随便乱抓几个都能看出区别来，但是和癌症的机理基本上就是打酱油的关系。
总体意思就是目前的主流方法找出来的marker不能作为推测机理的起点。
他们在faculty 1000上抱怨说他们这篇文章一共投了4年，前后被六个
顶级杂志据了（他们直接点名的几家估计是对他们态度最糟糕的，哈哈哈）
这里是他们在PLoS Comp Biol 发表的文章的摘要。
Bridging the gap between animal or in vitro models and human disease is
essential in medical research. Researchers often suggest that a biological
mechanism is relevant to human cancer from the statistical association of a
gene expression marker (a signature) of this mechanism, that was d... 阅读全帖 |
|
v*****s
发帖数: 20290 | 14 我知道的这个方法就是测分子在溶液里面的diffusion constant，还是make sense的。
你这么想两个极端情况，分子完全不动，那你的time trace要么是0，要么是1，反正是
个constant，time correlation是1，分子动得特别快，time trace都是从0一下子跳到
1，或者从1一下子跳到0，time correlation是0。其他时候介于两者之间，分子扩散越
快，time correlation越小。其中还有一个参数是分子体积，如果分子体积已知，就可
以算diffusion constant，反过来如果diffusion constant已知，就可以算分子体积。
不知道怎么测细胞。 |
|
v*****s
发帖数: 20290 | 15 我知道的这个方法就是测分子在溶液里面的diffusion constant，还是make sense的。
你这么想两个极端情况，分子完全不动，那你的time trace要么是0，要么是1，反正是
个constant，time correlation是1，分子动得特别快，time trace都是从0一下子跳到
1，或者从1一下子跳到0，time correlation是0。其他时候介于两者之间，分子扩散越
快，time correlation越小。其中还有一个参数是分子体积，如果分子体积已知，就可
以算diffusion constant，反过来如果diffusion constant已知，就可以算分子体积。
不知道怎么测细胞。 |
|
l**********1
发帖数: 5204 | 16 Sure
plus
one Australian University of Adelaide 2010 PhD Thesis
Title: Bayesian networks for high-dimensional data with complex mean
structure.
by Kasza, JE
Abstract: In a microarray experiment, it is expected that there will be
correlations between the expression levels of different genes under study.
These correlation structures are of great interest from both biological and
statistical points of view. From a biological perspective, the
identification of correlation structures can le... 阅读全帖 |
|
y*****w
发帖数: 1350 | 17 这篇文章在搞统计的看来就是一坨屎
娘希匹，Number of Nobel Laureates per 10 Million Population必然不是normally
distributed，而总国家数又不足30，不能用central limit theorem，所以那linear
correlation就是个意淫，只能是Spearman correlation，就甭谈用什么linear
regression做prediction了。而且NEJM这个档次的文章只做一个bivariate
correlation，不是侮辱读者的智商么。 |
|
E******T
发帖数: 59 | 18 Gene module在生物医学癌症分类（clustering），以及生物活性marker鉴定的应用
生物医学样品的分类（clustering）及其复杂，有几个原因： 医学样品的构成，比如
说白血病病人，有年龄，性别，癌症分级（I，II，III），用药情况，癌症类别（AML
，ALL）等等。按照不同的标准，就可以把病人样品分成不同的类别（clustering）.
更深一层次，不同的类别如果从生物学上来看，是由不同的基因，信号通路引起。如果
能找到这些不同类别对应的pathway,那么相对应的分类也就能被发现。 比如，白血病
里面的一种AML，他相应的信号通路就不同于另一种ALL，所以根据这些基因就能把白血
病分成AML和ALL。同理，如果能发现与癌症分级不同的信号通路，就能把白血病分成I
，II，III等不同的级别。但是，在平常的研究当中，这些具体的分类都不是特别清楚
。大多数情况下，仅仅知道其中的一种，比如在白血病里面就知道AML和ALL的分类，至
于其他的信息，很难得到。所以我们用了unsupervised learning的思想来研究这个问
题。
从生物角度来看，如果一个pat... 阅读全帖 |
|
l********n
发帖数: 260 | 19 Biomarkers. 2013 Feb;18(1):51-4. doi: 10.3109/1354750X.2012.730551. Epub
2012 Oct 16.
Serum fetuin-A levels are inversely associated with clinical severity in
patients with primary knee osteoarthritis.
Xiao J1, Wang XR, Hu KZ, Li MQ, Chen JW, Ma T, Li ZC.
Author information
Abstract
The present study aimed to detect serum fetuin-A levels in knee
osteoarthritis (OA) patients and to investigate their correlation with
clinical severity. We enrolled 215 knee OA patients and 76 healthy controls.
We m... 阅读全帖 |
|
t*****z
发帖数: 1598 | 20 生信学家的工作就是高屋建瓴，一览无余，先找到correlation，如果有条件，就挑几
个p-value最小的，安排些个实验员去试一试。
就好像司马懿观察武侯八卦阵，自信已经洞察其变化，遂唤出三员偏将，教他们从A门
杀入，从B门杀出，再从C门杀入，便可破这八卦阵。
: 了解correlation很重要，但不是终极目的。过度的沉浸于寻找correlation而忽
视本身
: 的物理事实只能把这个学科引向一条错路。这里又会有不少人会反驳说我们研究
的是生
: 物又不是物理。可是，生物研究的基本对象，像生化反应和分子相互作用，哪一
个又不
: 是一个个具体的物理过程呢？现在刚启动的human cell atlas项目，毫无疑问又
会给学
: 生信的兄弟们带来巨大的机遇，但是最终能在多大程度上促进我们了解生物过程
，我觉
: 得还很难说。用天文学来打个比方，是数天上有多少颗星星重要呢还是理解像万
有引力
: 定律这些普遍遵守的规则重要呢？
|
|
m*****n
发帖数: 3575 | 21 “不知道你是怎么证明经络存在的。你不可能通过解剖发现了经络的存在，除非经络比
细胞
核还小。如果你按压或针灸了某个经络部位，产生了疗效。这并不证明经络存在，因为可
能有别的原因。
证明人体中存在某一种物质（尤其像经络这种大系统）是一件非常重大的事情。非常重大
的事情需要非常客观的、直接的证据。”
——直接的证据就是刮痧显像。如果得某种病的病人主要在某几个穴位能刮出痧这么明
显、直接的证据都不能让你满意
你连照片都不用相信了。因为照片靠的是感光显像。
“coincidence多，correlation少，causality更少。coincidence里只有一
小部分是causality。
揉脾，然后头皮屑消失可能只是 coincidence，不一定是 correlation，更不
一定是 causality。
西医通过coincidence发现方法（当然通过其他的途径），通过临床试验研究
correlation，最终目的是证明有没有causality。”
——关于causality,中医自己有自己的中间步骤证明，其证明方法不依赖精密的机器，
而依赖气功炼的好的个人通过内感气流动而实现。如 |
|
f*****l
发帖数: 82 | 22 【 以下文字转载自 Statistics 讨论区 】
发信人: flywill (flywill), 信区: Statistics
标 题: Ask questions on Latin Hypercube sampling
发信站: BBS 未名空间站 (Sun Oct 7 15:17:10 2007), 转信
Is there anyone here who are familiar with Latin Hypercube sampling (LHS
)? It seems not easy to generate samples for correlated input. I mean th
at the elements in each input vector sample might be correlated.
It seems that there are 1 or 2 papers on correlation-induced LHS. Can yo
u suggest some?
Also, can I use the same method to const |
|
k******r
发帖数: 28 | 23 wooldridge本科教材和研究生教材中有这样两点：
1.time-series data情况下，error term Ut只要不与当期Xt相关(即zero
contemporaneous correlation)，即使与其他期Xt相关，OLS estimator也是
consistent的。
2.panel data情况下，每个individual有一个unobserved fixed effect Ct，Ct与Xt是
可以有arbitrary的correlation的。这种时候，fixed effect estimator要想
consistent的话，zero contemporaneous correlation就不够了，必须要满足strong
exogeneity，也就是Ut与所有期的Xt都不能相关。
想问问intuition是为什么？为什么有了unobserved fixed effect之后要求满足的
条件就强了？ |
|
b**********8
发帖数: 16 | 24 1.(1) Some Ph.D students proved that stock returns are auto-correlated. Is
this a violation of market efficiency?
(2) Assuming CAPM, and auto correlation, is it a violation or not? if it
is not, why? If it is, how can investors get the abnormal return.
Answer:
(1) Market efficient hypothesis states that when the market is in the
weak form efficiency, the market price has fully reflected historical
information. In other words, in an efficient market, the correlation of
stock returns is expec |
|
f*******r
发帖数: 257 | 25 When you have significant results, even though you have highly correlated
regressors, most likely you have a large data set. Although you have
significant results, the question you should ask yourself is: given x1 and
x2 have a correlation of 0.95, do they really both need to be in the model?
If your theory says yes, then it's fine. But if x1 and x2 have a
correlation of .95, they are very likely to be measuring the same construct. |
|
F****r
发帖数: 345 | 26 My understanding is you need to run a regression of A on B to find the
relationship between A and B.
Think about this: the stock price of Microsoft (A) probably will affect
market index (Z); the stock price of Coca-Cola (B) will also affect market
index (Z). Yet the stock price of Microsoft and Coca-Cola are unlikely
correlated.
Or think it this way: you may decompose Z into two parts, Z1 and Z2. Z1 is
correlated with A; Z2 is correlated with B. If Z1 and Z2 have no overlap (in
information), A a |
|
a*****g
发帖数: 19398 | 27 前一段，我把“围棋数学”的构思和进展情况介绍给了一位教授。
这位教授已经做了二十多年数学和工程的教育工作。
教授今天给回信——最后一句太鼓舞人心了——翻译成中文就是——
“如果你解决了2个难题，围棋数学将会席卷美国的每一个学校……”。
【 I do think that you are correct in noticing a correlation between math an
d go. Certainly, basic arithmetic... adding, subtracting, multiplying and th
eir geometric analogues are all there at the lowest levels of the game. But
also, estimating is there. And that is more abstract. Then the bigger, and
more interesting, question is whether or not the planning that comes up in
G... 阅读全帖 |
|
X*****e
发帖数: 64 | 28 按照我的理解，CWT 也是离散的，如果你看matlab的源码，其实也是产生一个波形来
correlate；但是CWT在dilation/translation上在“every possible scale ”上
correlate，这样计算量就大，其实也没必要。DWT只在dyadic的scale/shift上
correlate，计算量小，而且有快速算法(Mallat's tree structure)，并且证明能够完
全重构信号，所以用DWT。其实CWT/DWT这种命名有点misleading. |
|
z*****a
发帖数: 17 | 29 如何用已知的correlation coefficient 产生随机数列，使得数列中的每两个数都服从这
个已知的correlation coefficient? 有人说 decompose correlation coefficeint
matric, then linear transform of random variables? 但是我不清楚如何做，请明白
的大虾多多指点，多谢 |
|
f*****l
发帖数: 82 | 30 【 以下文字转载自 Statistics 讨论区 】
发信人: flywill (flywill), 信区: Statistics
标 题: Ask questions on Latin Hypercube sampling
发信站: BBS 未名空间站 (Sun Oct 7 15:17:10 2007), 转信
Is there anyone here who are familiar with Latin Hypercube sampling (LHS
)? It seems not easy to generate samples for correlated input. I mean th
at the elements in each input vector sample might be correlated.
It seems that there are 1 or 2 papers on correlation-induced LHS. Can yo
u suggest some?
Also, can I use the same method to const |
|
s****r
发帖数: 47 | 31 有三个联合高斯的随机变量X,Y,Z,不妨假定它们都是零均值并且方差相同,已知X和Y的
correlation coefficient是r1, Y和Z的correlation coefficient是r2,而且X->Y->Z形
成一个马尔可夫链,那么X和Z的correlation coefficient是多少.
答案是r1r2.
我想问的是有没有什么比较方便的办法得到这个答案.答案这么简单,我期待有一个比较
一目了然的解法. |
|
J*******g
发帖数: 267 | 32 I think at most an upper bound and a lower bound could be found for corr(x,
z). Looking at the correlation matrix of the random vector (x, y, z), which
would be a valid correlation matrix iff it is positive semidefinite. Note
that the only missing number in the correlation matrix is corr(x, z). Use
the positive semidefinite property, you can derive the bounds on corr(x, z). |
|
a*****g
发帖数: 19398 | 33 前一段，我把“围棋数学”的构思和进展情况介绍给了一位教授。
这位教授已经做了二十多年数学和工程的教育工作。
教授今天给回信——最后一句太鼓舞人心了——翻译成中文就是——
“如果你解决了2个难题，围棋数学将会席卷美国的每一个学校……”。
【 I do think that you are correct in noticing a correlation between math an
d go. Certainly, basic arithmetic... adding, subtracting, multiplying and th
eir geometric analogues are all there at the lowest levels of the game. But
also, estimating is there. And that is more abstract. Then the bigger, and
more interesting, question is whether or not the planning that comes up in
G... 阅读全帖 |
|
d**o
发帖数: 618 | 34 I respect Sandy Weill. Do you know how generous he supported Cornell's
School of Medicine?
Back to Dr. Carlos, discussing politics at office and not focusing on
patient care are very good reasons for firing. The main problem is that Dr.
Carlos did not focus on medicine. If he did research on epidemiology,
ideally with an MPH degree, then he should rely on his previous experience
and do quantitative comparisons between the two incidents. At the MPH
scholarship level, he should at least compare th... 阅读全帖 |
|
a*****8
发帖数: 2115 | 35 that's why I said it is controversial now. Originally, IBW is used even for
obese patients, however, studies have found that using IBW significantly
underestimate the creatine clearance where using TBW overestimate the
creatine clearance. alazar-Corcoran Equation is also used in some place that
factor in
height as well. Several study evaluated which equation best correlate with
creatine clearance and it appear that using lean body weight to calculate
eCrCl best correlate actual creatine clearanc... 阅读全帖 |
|
x******i
发帖数: 3022 | 36
Here's a standard procedure that works for all translation invariant
correlation functions.
1. Fourier transform correlation function to get spectral density.
[In your case, it is probably 8*pi*a^3/(1+a^2*k^2)^2]
2. Generate random data with delta-function correlation
3. Spatial Fourier Transform that data
4. Filter with Sqrt[spectral density]
5. Inverse Fourier Transform |
|
e****d
发帖数: 44 | 37 Quoted from an introductory book.
Correlational vs. experimental research.
Most empirical research belongs clearly to one of those two general
categories. In correlational research we do not (or at least try not to)
influence any variables but only measure them and look for relations
(correlations) between some set of variables, such as blood pressure and
cholesterol level. In experimental research, we manipulate some variables and
then measure the effects of this manipulation on other variables |
|
r******s
发帖数: 2155 | 38 【 以下文字转载自 Business 讨论区,原文如下 】
发信人: pringle (softbutcrisp), 信区: Business
标 题: A statistic question?
发信站: Unknown Space - 未名空间 (Wed Apr 27 18:59:39 2005) WWW-POST
If simple correlation shows that two variables are not significantly
correlated, could the two variables still be statistically associated?
Especially, could the nonsignificant simple correlation be caused by other
confounding factors?
Is it possible that by controlling for the other potential confounding
factors, we could still find th |
|
r*****t
发帖数: 286 | 39 ☆─────────────────────────────────────☆
kelvin527 (风云) 于 (Mon Mar 5 09:07:55 2007) 提到:
1. 三个变量，A和B correlation是0.9，B和C的correlation是0.8，问A和C的
correlation最
小是多少。
2. 两个点相距1m，A点波动率o1,B点波动率02，你在距A点xm处，问你感受的震动多少。
3. 8个球，分别写着1-8个数，放到8个盒子里，2和3能凑在一起的概率是多少？
☆─────────────────────────────────────☆
maglion (da木头) 于 (Mon Mar 5 10:01:36 2007) 提到:
1 cos(arccos(0.9)+arccos(0.8))~=0.46
2 what is 波动率? Frequencies and amplitudes are needed
3 2*7!/8! = 25%
少。
☆─────────────────────────────────────☆ |
|
c******e
发帖数: 101 | 40 If X, Y, and Z are 3 random variables such that X and Y are 90% correlated,
Y and Z are 80% correlated, what is the minimum and maximum correlation that
X and Z can have?
Dam hard |
|
b***k
发帖数: 2673 | 41 ☆─────────────────────────────────────☆
whatever1234 (哈哈) 于 (Sun Jun 22 01:54:24 2008) 提到:
Let X and Y be two gaussian random variables N(0, ) and N(0, ). X and
Y are correlated with a correlation .
What is the law of E (X − Y |2X + Y )
If we calculate the mean and variance of E(x-y|2x+y), it seems too tedious.
Any better ideas? 大牛
☆─────────────────────────────────────☆
bladehaze (^_^) 于 (Sun Jun 22 10:23:41 2008) 提到:
2 normal distribution with correlation, there is a formular |
|
b***k
发帖数: 2673 | 42 ☆─────────────────────────────────────☆
Flyinglan (flying) 于 (Mon Oct 1 22:22:31 2007) 提到:
三个RV X Y Z, X和Y的correlation是0.5， Y和Z的correlation是0.6，请问X跟Z的cor
relation的最小值？
个人感觉好像是个三角形之和等与180度，可能不对。
☆─────────────────────────────────────☆
robustzgy (浪迹天涯) 于 (Mon Oct 1 22:33:54 2007) 提到:
correlation is the cosine of the angle between two variables
angle 1 = arccos 0.5 = 60
angle 2 = arccos 0.6
so min(corr(x,z)) = cos (60+angle 2)
cor
☆─────────────────────────────────────☆
Flyinglan (f |
|
s***e
发帖数: 267 | 43 画图对于3个RV相对容易理解，对于多个可能不是那么直观了。我的理解是：
sample correlation converge to rho, so just consider sample correlations of
3 RVs. When each sample vector of
norm 1, then sample correlation is essentially the cosine of their angle.
The maximum angle you can get for 3 points on a unit sphere is 120 degree,
so you just compute the range
of cosine(alpha), alpha from 0 - 120 degrees.
The algebraic way of computing the determinant is in fact quite easy, if you
find out the right way. The
general formula for n RV |
|
s*****i
发帖数: 93 | 44 Three random variables: X, Y, and Z. If the correlation between X and Y is 0
.5, the correlation between Y and Z is 0.6. What is the range of the
correlation between X and Z？ |
|
t*******e
发帖数: 172 | 45 I am doubt about that
there is an example tell us, X, Y are not independenet N(0,1) with
correlation 0.
If your statement is true, then correlation 0 => independent? Since you
assume there are jointly normal which correlation 0=>independent.
The example is on Page 82 of Shreve's volume 2.
If jointly normal, nothing difficult about calculus, just P(X<=a,Y<=a), and
then take derivate of a, you do not need do tedious calculus. |
|
A********a
发帖数: 133 | 46 I guess u mean returns are auto-correlated.
One possible scenario is due to illiquidity, for example, some stocks may
not be actively traded, new information about it may take a few hours (days,
months) to realize those new info, the return series of such stock may have
positive correlation.
Another possibility is managers of the firms may massage profits/returns to make them more smooth.
Both illiquidity and return smoothness contribute to positive auto-
correlation in hedge fund returns, See C |
|
l********e
发帖数: 220 | 47 statistical arbitrage funds not necessarily have zero beta (although some of
them may claim to be, or want to be). But in general, they are expected to
have a low beta, or low market exposure.
Even some so-called "market-neutral hedge fund index" has significant market
correlation, for example:
http://www.hennesseegroup.com/indices/returns/year/2012.html
search "Market Neutral Index".. you can also check the previous years..
per my experience, a well-designed actual beta-neutral statarb portfoli... 阅读全帖 |
|
l*******1
发帖数: 113 | 48 Intuitively, when correlation is highest, its lowest
when the two are perfectly correlated, 1/3 chance A happens first.
When the two are negatively correlated, 1/2 chance A happens first.
I know nothing about credit. just intuition |
|
a*********a
发帖数: 3656 | 49 back when I worked in structured credit, correlation trading was a part of
the structured credit trading. correlation trading did index tranches, while
other groups in the structured credit did synthetic bespokes CDOs, cash
CDOs etc.
about the most exotic a correlation desk traded back then were those
leveraged super senior index tranches, which later turned into huge
disasters to them.
All mortgage products belonged to the mortgage desk. but I heard some of
them refer to themselves as structure... 阅读全帖 |
|
