l****z
发帖数: 29846 | 1 【 以下文字转载自 USANews 讨论区 】
发信人: lczlcz (lcz), 信区: USANews
标 题: Big Shot Investors Say No to IFRS
发信站: BBS 未名空间站 (Wed Nov 16 10:56:42 2011, 美东)
By Emily Chasan of WSJ
When SEC Chairman Mary Schapiro said in June that investors aren’t
clamoring for International Financial Reporting Standards, she may have been
understating things… a bit. Now, some of the biggest U.S. investor groups
are letting the SEC know in no uncertain terms that it should postpone its
decision on IFRS and even stop the converge... 阅读全帖 |
|
z*t
发帖数: 863 | 2 really great argument!!
我想morphorlogy上的convergent evolution时要不要考虑起始物种的问题，比如鲸的
祖宗在海边找食吃时，它们的genome，或者说molecular pool在就和鱼类有了相当大的
不同，但在长时间的evolution中，突变不断积累，根据morphorlogy phentype，产生
不同的适应性的progeny，可能一支变成河马，一支变成鲸。这种变化的molecular
base就可能和鱼类进化时完全不同。
这种百万年/上万个generation的molecular change，因为没有ancestor存在，基于进
化分析不能给出这种change的profile。
而且convergent evolution/divergent evolution现象不少和物种灭绝/环境变化相关
，这种自然状态下的mutagenesis更不可能用目前的molecular biology解释
可以观察的evolution估计就是岛屿上的达尔文雀了
directional does not stand”
西所以是有方向性的。我... 阅读全帖 |
|
d*****r
发帖数: 2583 | 3 发育的高度重现性也正说明有牢固的法则在指导这个过程
agree with this...
the embryo starts from just a few cells, and become such a complex system.
This means at least several things:
1. the initial states is limited (I assume all the initial information is
stored above or equal to the molecular level),
2. from a few cells to a whole organism, not sure if the cells contains all
information for the organism, seems likely; brain maybe different.
3. the final states is converged. Every human has two eyes and one nose. It'
s a... 阅读全帖 |
|
H*g
发帖数: 2333 | 4 I am not studying evolutionary biology neither. Just my 2cents. I guess I
would forget the neighbor-joining method.
Assuming that you have a good alignment with the most biological sense (
functional domain well aligned), good bootstrap value for ML and good
convergence for Bayesian Inference, regarding of the trees that are
generated from those different methods, you may have to distinguish them by
yourself by figuring out which one makes the most biological sense.
Without a thorough look at th... 阅读全帖 |
|
x*y
发帖数: 364 | 5 Help for reference of the following info.
Bilinear basis function (for 4-node quadrilateral element) yields higher
convergence rate than strict linear basis function (for 3-node triangular
element).
I can't find any book provides the info of the convergence rate for comparing
the linear and bilinear basis function. Could anybody by any chance know this?
Thanks! |
|
x*y
发帖数: 364 | 6 Now I'm also wondering about this.
I got my results that bilinear basis function (quadrilateral element in 2-D)
yields 1.89 convergence rate for conformation tensor (or elastic stress) and
linear basis function (tetrahedral element in 3-D) yields 1.65 convergence
rate for conformation tensor. I use DEVSS-SUPG method for viscoelastic flow
confpuation. A reference paper shows that DEVSS-SUPG method yields r+0.5 which
is 1.5 in my case. But I think 1.89 is far above 1.5, and my advisor thinks
bilin |
|
s**i
发帖数: 381 | 7 For example, one matrix is 2,170,159 X 2,170,159
it has 58,400,311 nonzeros
using preconditioned conjugate gradient method
it converges in 7 iterations.
If I use 32 processes, it only takes 0.28 seconds.
If I use 2 processes, it takes 5.29 seconds
The serial version converges in 4 iterations, taking 7.7 seconds. |
|
c*u
发帖数: 916 | 8 Which statistical method do you mean?
I just did a try. It looks like iteration works pretty well. Just need make
a guess of y_high and y_low, find the best thr. Then set y_high, y_low as
the mean values for xthr, find the thr again. Repeat the
iterations until converge. Only takes 2-5 iterations to converge.
Calculation
time is O(n) instead of O(n^3). For my data, it seems the fitting rarely
stuck in local maxima.
Sorry for the dumb question. |
|
i******t
发帖数: 370 | 9 Check the condition number of your Hessian first. If your Hessian is ill-
conditioned in the last steps, the convergence could be bad.
Also Newton's method could fail to converge if starting too far away from
the optimum. |
|
f******k
发帖数: 297 | 10 sorry, bad memory. that one proves NE converges to CE in a Cournot game
under some conditions.
The one that I wanted to mention is Roberts, Donald John & Postlewaite,
Andrew, 1976. "The Incentives for Price-Taking Behavior in Large Exchange
Economies," Econometrica, Econometric Society, vol. 44(1), pages 115-27,
January.
basically if the action set is the demand function, then usually you do not
get the convergence.
cannot
allocation
the |
|
G****n
发帖数: 145 | 11 He H 100 + papers
Convergence from discrete-to continuous-time contingent claims prices[PDF]
from yale.eduH He - Review of Financial Studies, 1990 - Soc Financial
Studies
Convergence of Contingent Claims Prices stocks and one bond available for
trading, markets
cannot be com- pleted by dynamic trading, and options cannot be priced by
arbitrage. This is
not the case in the continuous-time model, in which markets can be completed
by ...
Cited by 174 - Related articles - Library Search - All 13 ver... 阅读全帖 |
|
G****n
发帖数: 145 | 12 Between trust and control: Developing confidence in partner cooperation in
alliances
[PDF] from cuny.edu…, BS Teng - Academy of management Review, 1998 - JSTOR
Strategic alliances have been recognized as arenas with potential for
opportunistic behavior
by partners. Hence, a firm needs to have an adequate level of confidence in
its partner's
cooperative behavior. In this article we examine the notion of confidence in
partner ...
Cited by 1749 - Related articles - BL Direct - All 9 versions
A reso... 阅读全帖 |
|
n****t
发帖数: 170 | 13 For the z transform, we need to state the region of convergence. Different
region of convergence gives you different expansion. For example, X(z) = 1/(
z-1)= z^(-1)/(1-z^(-1)). If the ROC is outside the unit circle, x(n)=u(n-1),
which is causal; if the ROC is inside the unit circle, it is noncausal.
Actually for most cases, you only need to take a look at the ROC to tell
whether it is causal or not. |
|
G***T
发帖数: 66 | 14 1. solution frequency decides wavelength therefore determine the initial
mesh size and total meshs in the simulation. What happened in HFSS is based
on your solution frequency, HFSS will generate a initial solution, which
includes the mesh, port solution and etc. based on that frequency.
2. After the inital solution, HFSS will automatically refine the mesh and
perform a full solution until convergence based on that single solution
frequency.
3. If your model can be convergented at your solution |
|
o**a
发帖数: 76 | 15
算
this is not a counter example
by definition, \prod (1+a_n) converges if and only if
finite number of a_n are -1, and the partial products formed by the
nonvanishing factors tend to a finite limit which is different from 0
所以我刚才给的反例是错的, partial products趋于0，这infinite products不能说是
converge的，原因大概就是你上面给出的：那相当于\sum a_n=-00 |
|
B****n
发帖数: 11290 | 16 yes, for matric space, closed set means every convergent sequence converges to
itself. |
|
m*********s
发帖数: 5 | 17 Let {u_n} and u be probability measures on R^N, and suppose the characteristic
function of u_n , \int_{R^N} {exp(i ) u_n(dx)} , converges to that of u
, \int_{R^N} {exp(i ) u(dx)} , as n goes to infinity, for all t in R^N.
Show that \int_{R^N} {exp(i ) u_n(dx)} converges to \int_{R^N} {exp(i
) u(dx)} uniformly on compacts.
Is this a standard problem? Any hint? Thanks a lot! |
|
b****t
发帖数: 114 | 18
Hi DrumMania,
I do not have any convexity assumption on the objective function. If its
convex, then I think subgradient method and guarantee the convergence of
search to the optimum.
Can I ask the question another way: the steepest descent search method can
guarantee the convergence for what type of obj functions? Continuous and
smooth, and unimodular?
Thanks again,
Beet |
|
l*****e
发帖数: 238 | 19 if it's a series, only convergent for $x=k\pi$
if it's a sequence, convergent for any $x\neq 0$ |
|
g******a
发帖数: 69 | 20 the series is conditionally convergent,
not absolutely convergent. |
|
e**********n
发帖数: 359 | 21 I guess sum bi and sum ai converge, which are given conditions. So inf{bi} =
inf{ai} = 0.
Under these conditions, this inequality is not true. Counter example:
a1 = 1, b1 = 11/10, x1 = 1, y1 = 8/5
a2 = 76/45, b2 = 11*a2/10, x2 = 1, y2 = 1/2
and the remaining numbers can be made arbitrarily small, and let their sums
not only converge but also
be arbitrarily small.
Obviously y1+y2>x1+x2, and all the conditions are satisfied. |
|
c*u
发帖数: 916 | 22 ai.. unfortunately i have no idea how to build neural network.
I just did a try. It looks like iteration works pretty well. Just need make
a guess of y_high and y_low, find the best thr. Then set y_high, y_low as
the mean values for xthr. Repeat the iterations until converge.
Only takes 2-5 iterations to converge. Calculation time is O(n) instead of
O(n^3). For the data, it seems the fitting rarely stuck in local maxima.
Sorry for the dumb question. |
|
a****a
发帖数: 98 | 23 Based on the law of large numbers, the procedure will converge to the answer.
I don't know how many samples you need -- maybe 1000 should be more than
enough.
Though the number of possible permutations is large, it's still finite.
Similar algorithms based on sampling can work in cases where the sample
space has infinite number of elements.
You can easily write a program and test for the convergence.
sampling
large, |
|
J*******4
发帖数: 110 | 24 题目是这样的：Prove that there exists a subsequence $\{n_k\}$ of $\N$ such
that $\{\cos n_k\}$ converges.
Alternate wording: Consider the sequence $\{\cos n\}$. Show that this
sequence has a convergent subsequence.
偶的初步想法是找到一个整数序列使其模2pi后为一个收敛数列，可是这个想法看起来
似乎是不可能的。请问版上的前辈，这个问题到底该如何证明，在下先行谢过了！ |
|
j**********7
发帖数: 4 | 25 An interview question I met.
"Consider a matrix $J$. every eigenvalue of $J$ has positive real part.
Now let us consider infinite product of (1-J/n). Do you think this
infinite product converges? and how about the converge rate?"
Is there anyone can give some advice. thanks! |
|
h**********c
发帖数: 4120 | 26 我写这个程序的本意就是能让计算机本科二三年级的学生就理解，并且不需要任何特别
package或算法就可以用c或java把matlab的script重写出来。
这个版在数学以及算法上的高手是很多的，办法肯定有更好的。 由于时间关系， 我没
法试验了。牛顿法以前写过，挺长时间了。
如果说数学上更好的办法，这个点是一个fixed point,能找到一个quadratic
convergence，那么这是最快的。
我的稿还是打算投一投试试，主要希望一些定理，概念能应用到离散的convex hull 中。
大家对我的观点还是比较宽容的，如果您愿意看，过些日子可以发给您。
我打算先试再说，如果您也是相关刊物的editor，我现在就可以给您发。
手上一大堆数据要处理，这个问题需要放一放。
我相信数学方法应该是去构造一个|f''(x)|=0的函数算这个fixed point，quadratic
convergence 估计是不能再快了， 解一个系统用 1e-7左右的精度，通常就三到四个
iteration. |
|
g****t
发帖数: 31659 | 27 我是工程师。很多年没写过文章了。
我写这个程序的本意就是能让计算机本科二三年级的学生就理解，并且不需要任何特别
package或算法就可以用c或java把matlab的script重写出来。
这个版在数学以及算法上的高手是很多的，办法肯定有更好的。 由于时间关系， 我没
法试验了。牛顿法以前写过，挺长时间了。
如果说数学上更好的办法，这个点是一个fixed point,能找到一个quadratic
convergence，那么这是最快的。
我的稿还是打算投一投试试，主要希望一些定理，概念能应用到离散的convex hull 中。
大家对我的观点还是比较宽容的，如果您愿意看，过些日子可以发给您。
我打算先试再说，如果您也是相关刊物的editor，我现在就可以给您发。
手上一大堆数据要处理，这个问题需要放一放。
我相信数学方法应该是去构造一个|f''(x)|=0的函数算这个fixed point，quadratic
convergence 估计是不能再快了， 解一个系统用 1e-7左右的精度，通常就三到四个
iteration. |
|
a***s
发帖数: 616 | 28 Under certain conditions, this may be true. Try Berry-Esseen Theorem, which
gives the rates of convergence of the cdf. From that, we see that F_n conver
ges to F uniformly. You may prove that f_n converges to f for almost all x i
n R (not necessary all x). But Berry-Esseen Theorem has some assumptions.
N) |
|
M*****d
发帖数: 100 | 29 By CLT, Sn/（csqrt（n））converges to N(0,1) in distribution, not in
probability.
It's easy to find counter examples satisfying convergence in distribution
but not in probability. |
|
c*****t
发帖数: 520 | 30 请教parabolic equation收敛到stationary solution 的条件。
Consider parabolic equation
u_t=Lu. (P)
If we know that the second order elliptic equation
Lu=0, (E)
has unique solution v(x), then v(x) is the unique stationary solution of (P).
(We let (P) and (E) have same boundary conditions.)
Under what condition we can say that the solution u(x,t) of (P) converges to
v(x), as t goes to infinity?
In general, if (E) has several solutions v_1(x),...,v_n(x),... 阅读全帖 |
|
n***p
发帖数: 7668 | 31 This paper is in fact pretty interesting, even though the conclusion
is apparently wrong and the author is idiotic to submit it, and it is
ridiculous for the journal to publish it -- if the editor doe snot have
a grudge against the author.
The sum of the terms that contain no 9 is convergent. That part in the
paper is correct. Here I give a brief description.
As an example, look at the terms with 3-digit denominators.
1/100 + 1/101 +...+ 1/188, totally 9^2=81 terms with each term <1/100;
1/200 +... 阅读全帖 |
|
n******t
发帖数: 189 | 32 class A content(two semester):
Set theory/fundamentals. Axiom of choice, measures, measure spaces, Borel/
Lebesgue measure, integration, fundamental convergence theorems, Riesz
representation.
Radon-Nikodym, Fubini theorems. C(X). Lp spaces (introduction to metric,
Banach, Hilbert spaces). Stone-Weierstrass theorem. Basic Fourier analysis.
Theory of differentiation.
class B content(two semester):
Probability spaces. Distributions/expectations of random variables. Basic
theorems of Lebesque theor... 阅读全帖 |
|
S*****i
发帖数: 386 | 33 Let { X_n } be random variables, C is a constant and F(x) is a continuous
function,
if X_n converges to C in Prob.
Can we also conclude that
F(X_n) converges to F(X) in Prob.?
Thanks! |
|
t******g
发帖数: 1136 | 34 被一个题给难住了：
证明：
if a_1 + a_2 + a_3+ ....+ a_n + .... is convergent, then
a_1 + a_2/2 + a_3/3+....+ a_n/n + .... is also convergent.
谢谢各位指点。 |
|
l*******G
发帖数: 1191 | 35 the series does not have to converge, the expansion needs to converge to be
meaningful. otherthan that, they r the same |
|
B****n
发帖数: 11290 | 36 條件不完整巴
diagonal sequence最多也只可能證明到countable個點的pointwise convergence
如果要整個函數sequence pointwise convergence至少需要函數的其他條件 |
|
n**h
发帖数: 22 | 37 The Cramer-Wold theorem states that if every fixed linear combination of d
random variables converges to a normal distribution, then the d variables
jointly converges to a multivariate normal distribution. Does this theorem
hold when the dimension d goes to infinity? Thanks. |
|
n**h
发帖数: 22 | 38 The Cramer-Wold theorem states that if every fixed linear combination of d
random variables converges to a normal distribution, then the d variables
jointly converges to a multivariate normal distribution. Does this theorem
hold when the dimension d goes to infinity? Thanks. |
|
q*****z
发帖数: 191 | 39 Hi,
I have been thinking about the question of how to use ratio test to estimate
the radius of convergence of a power series when it only has even or odd
terms. For example, I have a complicated power series f(x):
f(x) = 1+f2*x^2+f4*x^4+...
To do a ratio test, estimate the following limit:
lim(fn/fn+2)
Then the value obtained is radio of convergence, but is it r or r^2? The
unit matches with r^2 but I haven't seen such things anywhere.
Thanks. |
|
n*********3
发帖数: 534 | 40 y=1+x+x^2+x^3+x^4+.... equation a
y=1+x(1+x+x^2+x^3+x^4+....)
y=1+xy
y=1/(1-x) equation b
now
equation a converges for when -1
but equation b converges for all x except when x=1.
How to understand or resolve this contradiction? which one is correct?
anyone?
many thanks in advance |
|
l***i
发帖数: 38 | 41 Hi every Ansys daxia,
I ran into a problem with using Ansys, I am checking my calculations with
Ansys on a thermal problem with radiation boundary condition. However, it
won't converge. Does anybody have experience on similar problem? I apprieciate
very much if daxia can suggest something on how to improve convergence in
nonlinear analysis.thanks a lot. |
|
S***p
发帖数: 19902 | 42 Prove of disprove the following claim
suppose Sum_{n=0,1...}f_n converges pointwisely to a function f on a set E
and {f_n} converges uniformly on E
then Sum_{n=0,1...}f_n uniformly on E |
|
e**********n
发帖数: 359 | 43 A figure definitely helps to understand my solution. See the attached file.
Each product xy is represented by a rectangle. As you divide the triangle
into more and more pieces, the sum of the area of blue rectangles converges
to the area of triangle ABC, as long as the maximum height of those
remaining small triangles converges to zero.
=
ADE |
|
j**********7
发帖数: 4 | 44 Consider a matrix $J$. every eigenvalue of $J$ has positive real part.
Now let us consider infinite product of (1-J/n). Does
infinite product converges? and how about the converge rate?
Is there anyone can give some advice? thanks! |
|
H****y
发帖数: 19 | 45 It's not (I-J/n)^n when n->infty, it should be Prod_n^infty (I-J/n)
If you take a ln, then the i^th diagonal element becomes Sum_n ln(1-Ji/n).
If you approximate ln(1-Ji/n) by -Ji/n, then it becomes Sum_n (-Ji/n), which
does not converge.
One can confirm this argument with Mathematica, using Ji=1, and calculate
Sum_n ln(1-1/n). You'll find the sum does not converge. |
|
p*****k
发帖数: 318 | 46 thanks. the question makes sense now.
but you sure it does not converge? seems to me you are saying the log of
this infinite product is going to -\infty, so the product converges to zero,
no? |
|
k*******d
发帖数: 1340 | 47 I agree. The Ito integral is well defined in L_2 convergence sense. To get
to the bottom of this, we need some math analysis. Google is not enough.
But I think for application, we do not care too much about this. Even in
Shreve's book, this convergence concept is explained in the footnote.
f(t)dB_t |
|
d*e
发帖数: 843 | 48 Thanks.
What I thought was using Martingale Convergence Theorem.
Since I(t) is L^2-bounded, it should converge both in L^2 and a.s. sense.
Since the L^2 limit is 0 (under conditions given in previous posts),
the a.s. limit must be 0 as well.
Correct me if I'm wrong. |
|
M****i
发帖数: 58 | 49 You are welcome.
Generally speaking, the martingale convergence theorem can not be applied
here because I(t) is only a semimartingale, not a martingale:
I(t)=g(t)*M(t),
where
g(t)=exp(-\int_0^t c(r)dr)
M(t)=\int_0^t c(s)*exp(\int_0^s c(r)dr) dB(s).
Note that g(t) is of finite variation and only M(t) could be a martingale.
This can also be verified by my previous example
c(t)=1/(1+t). In this case I(t)=B(t)/(1+t), which is not a martingale, but
it converges to 0 a.s. by iterated logarithm.
For yo... 阅读全帖 |
|
k***n
发帖数: 997 | 50 查了一下，是书上例题
\int_-inf^inf \frac{sin x}{x} converges not absolute
\int_c^inf \frac{cos x}{x} converges not absolute, where c>0
c <= 0 diverges. |
|
