|
c*****a
发帖数: 808 | 2 我以前看过nonparametric 的书，里面有介绍
Practical Nonparametric Statistics (Wiley Series in Probability and
Statistics)
我记得大概流程是有个s(x)的empirical distr function on x1,x2,...xN
然后建立一个null hypothesis some distr function F*(x). you are comparing x1
,x2...xN with F*(x) to see if it is reasonable to say F*(x) is true distr
function of the random sample.
I think the method is rather simple.
F*(x) is usually starting with 0 to 1 , by .01 .05... etc your choice based
on size of n. You usually have to make a chart ... 阅读全帖 |
|
发帖数: 1 | 3 我是F2大妈，靠探亲出国，如今在廊坊靠发帖为生，一贴5毛
你跟我讨论这个几何分布，是不是太抬举我了
我知道的很有限， poisson distr, uniform distri, binomial distri, bernuli
distr
就这点水了，哈哈哈 |
|
l******e
发帖数: 470 | 4 I don't think this is a good test.
just by a similar argument of coupon collector problem, only 1-1/e grid will
be nonempty if the distr is truly uniform. One the other hand a lot of other
weird distr will behave the same way.
One way is to throw NlogN point and test whether each grid has (1+-\epsilon)
logN
points, If it is uniform ,then this should happen w.h.p.
However, I don't this is standing on a solid ground neither. you need
Goodness-of-Fit Test type of stuff.
)x
than
some
making |
|
m***8
发帖数: 8 | 5 第二个部分可不可以这样算 也是接着第一部分的想法 四个角的叫做A类 八个靠边的叫
做B类 四个中心的叫做C类
不同类之间的transition prob应该好算
A->B: 1
B->C: 1/3
B->B: 1/3
B->A: 1/3
C->B: 1/2
C->A: 1/2
然后得出stationary distr.
然后相同类的格子应该是平均分布 继而得出整个的stationary distr. |
|
m***8
发帖数: 8 | 6 得到一只小老鼠情况下的4*4的棋盘中的stationary distr之后
你应该问自己的问题是
如果有16个小老鼠在这个4*4的棋盘 那么空格个数的distribution的个数要怎么算
这个既然已经是stationary的
那么这16个小老鼠就是被一个一个
以stationary distr被扔在棋盘上
这个可以这样算出来 |
|
b******3
发帖数: 4385 | 7 近5年有很大希望打入国际市场，大家应该去看看这个distr 的排名！出自湖北一个很
小的民营企业！这种事情拿很多公家经费的政府科研机构其实应该自觉点，不能再骗钱
了，人家5，6个人的小团队，一分经费不要，做得比中科院得好 |
|
d*********2
发帖数: 48111 | 8 毕业不难， 给老黑的分也是有加成的。
如果按standard distr. 我带过的老黑基本上都是最后的几个， 但是每次分数下来，
人往往还能拿C+, 在２５人的班里大概是１８－２２名的样子。 |
|
B*Z
发帖数: 7062 | 9 别喷了。audi 55:45是suv，和丰田的highlander基本一样的。audi a4的weight distr
ibution和camry差不多。 |
|
l*****6
发帖数: 7881 | 10 你再查查。我查到的audi a4 B8平台的也是55:45
distr |
|
C*****z
发帖数: 2050 | 11 也看是哪年哪款了，比如上代Q7就是大约50:50，这代Q7就接近55:45了。最近这两代
Camry还是更接近60:40。
distr |
|
D***I
发帖数: 1957 | 12 别搞笑了
爬坡王跑不上坡最大问题就是前面太轻
distr |
|
r***e
发帖数: 2539 | 13 从节省tax的角度考虑，我有一点不大清楚：
假设我在普通的brokerage account有一个共同基金， 每个月有固定的divident distr
ibution，会不会触发tax呢?
如果会的话，是不是在Roth账户买这些基金比较合适呢？在普通账户放些没有分红的基
金？
先谢谢了！ |
|
T*********e
发帖数: 9208 | 14 yes to all ur questions.
distr |
|
E*******0
发帖数: 465 | 15 1. rand5()->rand2() and rand4();
2. we can find 11/12/13/14/21/22/23/24, and we can find uniform 12/13/14/21/
22/23/24.
for 1. If we have 5 balls (1,2,3,4,5), the prob of choosing each ball is 1/5
. If we only choose 3,4,5, the prob of choosing each ball (one of 3,4,5) is
1/3. So we can easily get rand2() and rand4().
for 2. since all the 8 cases are uniformly distributed, when we choose the
first ball as 1 and choose the second ball as 1, we just ignore it.
So we can get 7 cased uniformly distr |
|
h***m
发帖数: 85 | 16 We are in process now, if you do not hear from us within 4 weeks, please move on. Thanks!
If you are interested, please send me your resume.
PS: our company sponsors H1-B(Non-profit)
Salary: $67,118 - $83,897
Position Description
The Data Analyst is responsible for ongoing analysis of migrant student
information, obtained from a variety of sources. The Data Analyst is
involved in the development and preparation of information products for
Federal, State, Migrant Regional Offices and school distr |
|
|
s***g
发帖数: 1250 | 18 想买那个100硬币的bag，这段话说得是100个相同的coin吗？如果这样，就没必要买了。
谁能帮我看看下面这段话1
The Virgin Islands quarter-dollar is the 5th coin released in the 2009 Distr
ict of Columbia and U.S. Territories Quarters Program. This bag is filled wi
th 100 Virgin Islands quarters struck on the main production floor of the Un
ited States Mint at Philadelphia, each bearing the “P” mint mark. |
|
w******e
发帖数: 609 | 19 谢谢。还有一个问题。
我是SOGOTRADE。在它的1099statement上
9. UNREALIZED PROFILE OR (LOSS) ON OPEN CONTRACTS -- 12/31/2008
10. UNREALIZED PROFILE OR (LOSS) ON OPEN CONTRACTS -- 12/31/2009
这俩项都是空白。
而在
1a. TOTAL ORDINARY DIVIDENDS 1.00
1b. QUALIFIED DIVIDENDS 1.00
2a. TOTAL CAPITAL GAIN DISTR. 0.00
在1040上说如果2.a是0.就不用报。是真的吗？ |
|
w******e
发帖数: 609 | 20 谢谢。还有一个问题。
我是SOGOTRADE。在它的1099statement上
9. UNREALIZED PROFILE OR (LOSS) ON OPEN CONTRACTS -- 12/31/2008
10. UNREALIZED PROFILE OR (LOSS) ON OPEN CONTRACTS -- 12/31/2009
这俩项都是空白。
而在
1a. TOTAL ORDINARY DIVIDENDS 1.00
1b. QUALIFIED DIVIDENDS 1.00
2a. TOTAL CAPITAL GAIN DISTR. 0.00
在1040上说如果2.a是0.就不用报。是真的吗？ |
|
i******s
发帖数: 925 | 21 借贴问一下，我的情况与楼上很像。我交易获得了，但1099-DIV的BOX2A是0
（2a. TOTAL CAPITAL GAIN DISTR---------0），看1040上line13说如果这种情况可以
不用SCHEDULE D，我不用补交这部分税吗？我的line13就填0吗？谢谢 |
|
w******e
发帖数: 609 | 22 谢谢。还有一个问题。
我是SOGOTRADE。在它的1099statement上
9. UNREALIZED PROFILE OR (LOSS) ON OPEN CONTRACTS -- 12/31/2008
10. UNREALIZED PROFILE OR (LOSS) ON OPEN CONTRACTS -- 12/31/2009
这俩项都是空白。
而在
1a. TOTAL ORDINARY DIVIDENDS 1.00
1b. QUALIFIED DIVIDENDS 1.00
2a. TOTAL CAPITAL GAIN DISTR. 0.00
在1040上说如果2.a是0.就不用报。是真的吗？ |
|
w******e
发帖数: 609 | 23 谢谢。还有一个问题。
我是SOGOTRADE。在它的1099statement上
9. UNREALIZED PROFILE OR (LOSS) ON OPEN CONTRACTS -- 12/31/2008
10. UNREALIZED PROFILE OR (LOSS) ON OPEN CONTRACTS -- 12/31/2009
这俩项都是空白。
而在
1a. TOTAL ORDINARY DIVIDENDS 1.00
1b. QUALIFIED DIVIDENDS 1.00
2a. TOTAL CAPITAL GAIN DISTR. 0.00
在1040上说如果2.a是0.就不用报。是真的吗？ |
|
x****9
发帖数: 756 | 24 subject categories COMPUTER SCIENCE, THEORY & METHODS:
IEEE T NEUR NET LEAR Rank 2 of 100
IEEE T PARALL DISTR Rank 16 of 100
subject categories COMPUTER SCIENCE, HARDWARE & ARCHITECTURE
IEEE T NEUR NET LEAR Rank 1 of 50
IEEE T COMPUT Rank 14 of 50 |
|
x****9
发帖数: 756 | 25 subject categories ENGINEERING, ELECTRICAL & ELECTRONIC:
IEEE T NEUR NET LEAR Rank 9 of 243
IEEE T PARALL DISTR Rank 64 of 243
IEEE T COMPUT Rank 92 of 243 |
|
l********r
发帖数: 28 | 26 看到大家要打官司告奥本，真的很激动。我代表LIA坚决支持大家的革命行动。另外，
我从robert那里要来了08年LIA法律小组对起诉奥本排期捆绑的可行性报告。大家看看
附件。
虽然这个报告的意见比较负面，但我们相信长江后浪拍前浪，前浪必死沙滩上。新PD们
一定有更好的法律分析能力，组织能力和宣传能力。也殷切的希望新PD能站在我们失败
的基础上，勇敢地闯出一条新路来。
如果需要帮助，请与我们联系
谢谢
I. Jurisdiction and Venue:
1. Court has subject matter jurisdiction over this action pursuant to 28 U.S
.C. § 1331 (federal question jurisdiction; and the Due Process Clause of
the United States Constitution. The plaintiffs bring their claims pursuant
to 8 U.S.C.§§701 et seq (Administrative Proced... 阅读全帖 |
|
|
g******4
发帖数: 6339 | 28 Let me finish my friend's story first.
He came to the states when he was six. His parents sent him to a private
school first (not expensive, about $200 a month), and taugh him math/reading
after school using the books from http://www.teachercreated.com/retailers/
His parents believed in Chicago manget schools, and wanted him study with
kids from different backgroud, specially those from low-income families. In
that Nov-Dec, his crazy parents sent out application to 15 magnet schools in
the distr |
|
|
|
j***y
发帖数: 882 | 31 life in financial distr is boring, esp in weekend. totally dead...
what can you do there for fun?
studio了 |
|
l***1
发帖数: 1223 | 32 hudson也分好几个学区吧？好像是hudson city school district 不错。有的地址是
hudson，可是school district怎么在twins burge school dist 或者stow-nunroe
fall city school distr? 有没有人能详细解释下？
还有，查了下top 10 ohio的school district，咋没看到大家都说最好的solon？还是
solon的school district叫另一个名字？
最后，在akron上班，值得买solon的房子每天跑30几英里吗？还是买hudson房子就可以
了？ |
|
|
g*******1
发帖数: 8758 | 34 我觉得无神论是linux，道教是个distr，呵呵 |
|
b*****l
发帖数: 9499 | 35 昨天和朋友聊了一下，发现很多人认为这个属于个人隐私。
我说一下我的观点哈：我觉得，当有 conflict of interest 时，大家对公共管理者的
宗教信仰有知情权；大家质询其宗教倾向本身并无不妥；公共管理者当然可以拒绝澄清
，但要承担由此引发的不信任；知悉者爆料，并不构成侵犯隐私。
而对普通 ID 来说，知情的爆料，算侵犯隐私；但就宗教相关话题讨论时，别人可以询
问，自己可以选择回答或不回答。
我觉得，钻风属于公共管理者；甚至，更进一步，BM 也属于。
再展开一些，任何版面，与版面相关的个人信息，大家如果询问，也是正当的。比如说
，申请 Linux BM 的，大家有权问其用过哪些 Linux Distr；申请 PhotoGear BM 的，
大家有权问其相机是哪家的；Wisdom BM，大家有权问是否信奉儒释道中的一门，具体哪
个学派或法门。当然，BM 有权拒绝回答。
大家怎么看？ |
|
T**e
发帖数: 424 | 36 Ada不也是一个很优秀的OO语言吗，不是也因为DOD
的封闭而成为实际上的DOD PL了？
Java目前的很大一个好处就是拥有stable, expanding
resources.可以找到的tools, packages很多，其
应用跨越standalone appl, distr. appl ( including
web appl.), embedded system,支持的平台很多，
拥有的programmer很多，这些财富都是目前其他PL无法
相比的。 |
|
L***n
发帖数: 6727 | 37 强烈同意, XP命令行界面其实很强，某些地方甚至比Linux还强，我看最方便的界面就
是尽量少用鼠标的图形界面，XP在用命令行控制图形界面方面可以做到很爽，比如切换
窗口用键盘搜索匹配窗口标题，工作起来如果开一堆窗口外加几个窗口放歌浏览什么这
个功能爽歪了，而且几行script就可以实现这个功能(with Autohotkey)。Win+R是
surprisingly good，好过我用过的任何Linux distr类似功能。在加上cygwin和
gnuwin32，现在的xp是我最满意的桌面平台。无论速度，用键盘控制的广泛程度，
界面的consistency还是script定制方面都强过ubuntu
户都 |
|
c*********3
发帖数: 197 | 38 Tried to install several different linux distr on an old IBM eServer. Only
debian works. |
|
mw
发帖数: 525 | 39 现在需要加一个远程的服务器，主要是SQL和一些python，科学计算，socket之类的应
用，应该用哪
个distribution比较好呢？
现在candidates包括
fedora7，8，9
debian, 4, 5
ubuntu 8, 9
现在我比较群像debian，fedora的distribution好像不太好，很多我看到的包都没有他
的rpm，
ubuntu好像说是有很多地方跟标准的linux不兼容，我担心万一将来有个什么lib不工作
我就s掉了
可是我碰巧从来没用过debian
想请各位大侠给点建议！ |
|
a*****i
发帖数: 4391 | 40
If used primarily for server (meaning you will not care about the fancy user
land utility programs), install CentOS. |
|
w********e
发帖数: 557 | 41
The bottom of line is none of the linux distros is good enough to become
main stream any time soon. All the "choices" are pretty much
meaningless since only geeks care about them. I wouldn't call ubuntu
that much a success. Its market share is insignificant.
Simply put, linux is too difficult to use, even it is free. People want
sth that more or less work out of the box. Linux is far and far away
from that. Part of the reason is all the development resources are
diluted into so many distr |
|
b***y
发帖数: 2799 | 42 ☆─────────────────────────────────────☆
slipper (ohh) 于 (Thu Oct 13 20:16:21 2005) 提到:
anybody know if VC++ has a function , which take a random number as input (the
random number is between 0 and 1), and it use inverse method to output a
value, which follow normal distri.? or dose VC has normal distribution random
number generator? Thanks.
☆─────────────────────────────────────☆
fightwater (QT) 于 (Fri Oct 14 12:53:06 2005) 提到:
using box-muller algorithm
http://en.wikipedia.org/wiki/Bo |
|
r****t
发帖数: 10904 | 43
DaShagen 的叙述是有些地方让我觉得不清楚，所以我把他的话重说了一遍，见前面帖子。我也没有说要用不带假设的test，我只是说 t test 不假设 underlying distribution is normal. 你没搞清楚我们在说什么，有时候你读贴仔细点，能省你自己和大家很多时间。
两均值之间的 t test 对 underlying distribution 没有 normalty 假设，只要均值
存在就行，i stand by this statement. 严重偏离 normal distribution, 只要是
iid sample. sample mean 仍然是 normal distribution, 仍然用 t test. 有时候你
只要读书认真点，能明白很多事。
均数仍然是存在的，仍然是有定义的，只是在这种情况下 t-statistic 没有“代表性
”，这个和没有"意义"是两码事。你要是追求“代表性”当然不能 t test. 但是即使
在这种情况下 t-test 仍然是有意义的，其结论 on means of the underlying
distr |
|
t****g
发帖数: 715 | 44 No big difference between logit and probit, however, you choose one of them
based on your preference. For instance, it may make your life easier to
choose probit, as its error is normal, which can either fit your model
better or make your presentation more reasonable for people outside
econometrics(you know, sometimes it is hard to explain to dumb guys why you
need a logistic distribution instead of normal). Historically, logit was
thought easier to compute during the period when computing distr |
|
a**********m
发帖数: 2098 | 45 you are way below the bar...
i think i can improve to 10 hours a day if i come to here less and not distr
acted by kaixin001 or something like that... |
|
o*********s
发帖数: 275 | 46 In my 1L year, I basically studied all the time, except when I was driving,
eating, sleeping, and taking care of my son. Now I hang around bbs all the
time.
distr |
|
W*********l
发帖数: 624 | 47
distr
but classes take, say, 8 hrs?
i might be able to do that when i was a highschool kid and family took care
of me.
now we are all grown ups, and have to take care of ourselves, which is very
distracting and time consuming. |
|
r*****t
发帖数: 286 | 48 ☆─────────────────────────────────────☆
theta (delta, gamma and vega) 于 (Tue May 1 22:59:25 2007) 提到:
(x1,x2,x3)--multiNorm(u,sigma), and we know u and sigma.
then condition on x3=x, what is the covariance matrix of muitinorm distr
ibution (x1,x2)?
anyone can shed light on this?
☆─────────────────────────────────────☆
quantler (quant) 于 (Tue May 1 23:03:41 2007) 提到:
熬了一个通宵后 我的神智已经不是很清醒了
不能帮您了：）
☆─────────────────────────────────────☆
theta (delta, gamma and vega) 于 (Tue May 1 2 |
|
k*******d
发帖数: 1340 | 49 X_n are N(0,1)
X_0 is -X_n
clearly X_n -> X_0 in distribution since they have the same distribution;
However
P(|X_n-X_0|>\eps) = P(|2X_0| > \eps), where X_0 is N(0,1), clearly this prob
does not go to zero.
还有，converge in distribution不要求随机变量在同一个probability space上，
converge in probablity 要求在同一个prob space上，除非X_0是常数
Note that if X_n -->(in distr) X_0, where P(X_0 = c) = 1, c is a constant
then X_n --> X_0 in prob |
|
