l**********r
发帖数: 12 | 1 sorry I don't understand
could you explain a bit how to connect argmin{Bt} with 3.3' in the paper?
Thanks |
|
y***s
发帖数: 23 | 2 Pr(min{B_t,0a)=Pr(min{-B_t,0a)(B_t and -B_t has the same
distribution)
=Pr(-max{B_t,0a)
=Pr(max{B_t,0
=1-Pr(max{B_t,0-a)
1-exp(-2a^2) by 3.3' in the paper
The pdf of distribution of argmin{ Bt, t in [0, 1] } is
(-4a)exp(-2a^2)
where a is negative
by taking derivative of 1-[1-exp(-2a^2)] with respect to a. |
|
k*******d
发帖数: 1340 | 3 My understanding is argmin{Bt, t in [0,1]} is the argument t, so we are
asked to find the distribution of t_min where B(t_min) is the mininum of B(t
) when t in [0,1]. It seems to me you are finding the pdf of the minimum
value. |
|
t****a
发帖数: 1212 | 4 这是个数学题诶
假定用XY = {[x1,y1],...,[xi,yi],...[xn,yn]} 表示各个人的位置
求[x0,y0] = argmin{\sum(f(x0,yo))} = argmin{\sum(dist([xi,yi],[x0,y0]))}
目标函数是关于x0,y0的二次函数，可以对它求关于x0,y0的偏导数，最小值发生在两个
偏导数都等于0
解这个方程组可以得到x0, y0的取值 |
|
b*****e
发帖数: 14299 | 5 ☆─────────────────────────────────────☆
skydive (跳跳~~修竹凝妆，垂杨驻马) 于 (Wed Aug 1 22:40:57 2012, 美东) 提到:
发信人: greendia (绿代), 信区: Investment
标 题: 我这种情况怎么理财？（请勿顶置，谢谢！！））
发信站: BBS 未名空间站 (Sun Jul 29 20:55:08 2012, 美东)
双职工分别毕业2年，1年。存款10万，绿。
两车，廉价condo学生时代已付清。没有小孩计划。
HSA 两年MAXOUT 1万6。
401K 0（公司没MATCH)
目前最大的开销是吃饭。
目前的钱存的是checking，0.8%左右的汇报。
关于这10万，请问怎么才算明智的选择：
1. 继续存house，争取现金付清（德州）
2. 在同一小区买一个condo出租，月租房价比例1.7%。
3. 买基金，炒股？（感觉不太有天赋)
请问哪个选择好些？感觉3对我们来说不太靠谱。
☆─────────────────────────────────────☆
sk... 阅读全帖 |
|
N**D
发帖数: 10322 | 6 \newcommand{\argmin}[1]{{\hbox{$\underset{#1}{\text{\textnormal{argmin}}}\;$
}}} |
|
T*******n
发帖数: 493 | 7 \documentclass{article}
\usepackage{amsmath}
\DeclareMathOperator*{\argmin}{argmin}
\begin{document}
In text $\argmin_{x\in X} \{j | g(X_j) \}$ and in displayed equation
\begin{equation}
\argmin_{x\in X} \{j | g(X_j) \}
\end{equation}
\end{document} |
|
发帖数: 1 | 8 试一下 一种方法是求点到空间直线最短距离点 。 空间直线：x=f(z)&y=g(z);. 点 （
x0 y0 z0)； 那么答案就是解 argmin(x0y0z0 到 xyz的距离） |
|
t****a
发帖数: 1212 | 9 来自主题: JobHunting版 - 请问G一题 给了个DP的解法，请各位指教
Define A={a_1, a_2, …, a_n} as the group of number, n is even
Let X={x_1, x_2, …, x_n}, x_i ∈ {-1,1}
Goal: argmin(Y=∑{A*X}), subject to ∑{X}=0
DP formula:
Y = Y_n(n, 0, 0) = min(Y_{n-1, A_n, -1}, Y_{n-1, -A_n, 1})
详细代码参见
http://kangtu.me/~kangtu/study/interview.html#sec-11
原文中的例子结果为 [-4 [[4 9] [1 16]]] ; 差为-4
该平方数列延长到10的结果 [-1 [[1 25 36 49 81] [4 9 16 64 100]]] ; 差为-1
延长到20: [0 [[1 49 81 100 121 144 169 225 256 289] [4 9 16 25 36 64 196 324
361 400]]] ; 差为0 |
|
t****a
发帖数: 1212 | 10 来自主题: JobHunting版 - 请问G一题 给了个DP的解法，请各位指教
Define A={a_1, a_2, …, a_n} as the group of number, n is even
Let X={x_1, x_2, …, x_n}, x_i ∈ {-1,1}
Goal: argmin(Y=∑{A*X}), subject to ∑{X}=0
DP formula:
Y = Y_n(n, 0, 0) = min(Y_{n-1, A_n, -1}, Y_{n-1, -A_n, 1})
详细代码参见
http://kangtu.me/~kangtu/study/interview.html#sec-11
原文中的例子结果为 [-4 [[4 9] [1 16]]] ; 差为-4
该平方数列延长到10的结果 [-1 [[1 25 36 49 81] [4 9 16 64 100]]] ; 差为-1
延长到20: [0 [[1 49 81 100 121 144 169 225 256 289] [4 9 16 25 36 64 196 324
361 400]]] ; 差为0 |
|
r*g
发帖数: 186 | 11 right,
这个是weiszfeld问题, 用fixed point可迭代
另外那个street问题求教
我的做法|x - xi| = ri 然后
argmin sum(ri)
st. x - xi + ri >= 0
x - xi - ri <= 0
转化成LP问题 我觉得有更方便的方法 |
|
c********y
发帖数: 30813 | 12 等你买了再说吧
不定到时候，又来做你的argmin了 |
|
m****s
发帖数: 402 | 13 Many thanks!
For argmin(a'Aa), A is known, how to get a from A? |
|
|
b****t
发帖数: 114 | 15 \[ argmin_{x \in X} \{j | g(X_j) \} \]\
I want the domain x \in X 在 argmin 正下方.
how to do that please? thanks.
Beet |
|
b*********n
发帖数: 173 | 16 Sorry. should be
\[ argmin\limits_{x \in X} \{j | g(X_j) \} \]\ |
|
b****t
发帖数: 114 | 17 Hi all,
I want to add some extra margins on either side.
like this
39 Consider the integer-optimization problem: \\
40 \\
41 \begin{tabbing}
42 $\mathbf{(P)}$ \= \kill
43 \> find $\underline{x}^*$ such that \\
44 \[ \underline{x}^* \in \argmin{\underline{x} \in \mathbb{X}}\: g (\
underline{x}) \]
45 \< where $\mathbb{X} \subseteq\mathbb{Z}^d$ and $\mathbb{Z}^d$ is
the set of $d$-dimensional
46 integer vectors, and $g:\mathbb{X} \to \mathbb{ |
|
s***n
发帖数: 9499 | 18 thanks:)
But it looks like I didn't state my question clearly. :)
argmin{tr((A'A)^(-1))}
A
Where 0<= Aij <=1.
Invertible also implies A is full rank.
Let A be a n by n matrix.
I know the result for n = 1, 3, 7, 11, ... (4k-1)
For example, when n = 3,
A =
[
0 1 1
1 0 1
1 1 0
],
which is related to Hadamard matrix for n+1 = 2, 4, 8, 12, ...(4k).
I would be appreciate if you can give some hints for other values of n such
as 4, 5, 6. |
|
l*****4
发帖数: 6 | 19 问题如下：
min a*x
s.t. A*x+argmin B*x>b
x
该如何用LP来解决？谢谢！ |
|
s*****c
发帖数: 753 | 20 right. So this problem is t=argmin(distance).
don't know how to find the minimum? derivative = 0?
Are you simply asking for a solution or are trying to solve it? |
|
n*******l
发帖数: 2911 | 21 Least square is to minimize the L2 norm of the errors, that is,
k=argmin{ (y1-k x1)^2 + (y2-k x2)^2 }.
这基本上是描述这两个点在 y方向上离这条直线的距离，跟楼主原文里的要求不是一回
事。 |
|
l**********r
发帖数: 12 | 22 how to calculate the distribution of the time of reaching minimum of a
Brownian Motion Bt? 0<=t<=1
I guess that its pdf looks like
- symmetric around 1/2
- higher at 0, 1 and lower at 1/2 |
|
|
l*****4
发帖数: 6 | 24 问题如下：
min a*x
s.t. A*x+argmin B*x>b
x
该如何用LP来解决？谢谢！ |
|
m*********n
发帖数: 413 | 25 是哦。
这题不会是这样子的吧
Y_outcome1 = ...
Y_outcome2 = ...
var(Y_outcome1) = ...
var(Y_outcome2) = ...
need (w1,w2) = argmin(var(w1*Y_outcome1+w2*Y_outcome2)) given w1+w2=1
这不就是选其中一个model嘛。。 |
|
h***i
发帖数: 3844 | 26 argmin sum (y1-a0-a1x1-a2x2)^2+sum(y2-a3-a1x2-a3x1)^2
a0,a1,a2,a3 |
|
