由买买提看人间百态

topics

全部话题 - 话题: variances
首页 上页 1 2 3 4 5 6 7 8 9 10 下页 末页 (共10页)
c*****m
发帖数: 4817
1
来自主题: Statistics版 - 请教: Var(Xi-X-bar) 等于Var(Xi) 吗?
你既然想计算xi-xbar的variance,这里xi就是作为一个随机变量。
第一个计算是对的,之前你混淆了variance和sample variance,后来改过来了。
第二个问题很大,你的xi是random variable,所以你才能计算它的variance,
但后面x1,x2 .. xn又变成了xi的sample,很混乱。

..
v********g
发帖数: 250
2
来自主题: Statistics版 - sample distribution的理解
你好,首先,这是一个binomial和sampling distribution综合问题。
第一,大体上看,这是一个sampling distribution的问题。也就是说,从population
里randomly select samples。 所以,这是标准的sampling distribution的问题,但
关键在于,需要知道 population mu 和 variance.
(x-u)/sqrt(variance) 就会是normal distribution了。
第二, 题目本身是 binomial的问题,也就说,红球出现follows binomial
distribution. 而红球出现的 mean = n*p , variance= n*p*(1-p), 而这两个就是之
前要求出的 population mu 和 variance。
最后,本题目要求的是 P(X〉40),把X 换算成 Z, 这样就能得出最后的结果了。
希望对你有帮助。
R*****0
发帖数: 146
3
To LZ, it seems IMO the process of drawing samples (either 10 or 8) is not
independent each other. An average correlation of 0.9 means you have a lot
of sample pairs with very high (close to 1) correlations, which is not
likely to happen if you draw samples randomly. Anyway, it still does not
tell you much about the sample variances.
For example, if you draw 8 samples from population 2. Now if you
multiply all those data by 2, you will get the same 28 correlations.
However, the variances will be... 阅读全帖
a*****i
发帖数: 1045
4
一个Mixed model的模型。两组人类,一组有病的,一组没有病,5个时间点,主要比较
他们的mmse score 随时间的变化,covariate 有年龄和housing conditions(3个不同
的housing conditions) 这这个model里面,我们用了random intercepts and random
slopes.
1. 在开始时间段,发现两组人类,两组的mmse score的variance 一组大,一组很
小,问问 mixed model 怎么take this into count, 还有两组observations,一组明显
很多人,一组很少人。想问问mixed model count for random effects, 那么人在这里
就是random effects, mixed model 是相当于两个个组分开做linear regression的,
所以并无所谓两组人,一组的variance很大,一组的variance很小。还有两组人数不一
样的问题,unbalance data都可以解决了吗?
2. 我们做了一个cor... 阅读全帖
t**********y
发帖数: 374
5
I did some test runs using built-in dataset and got different outputs:
1. ===
> test=lmer(Reaction~Days+(Days|Subject),sleepstudy)
> summary(test)
Linear mixed model fit by REML ['lmerMod']
Formula: Reaction ~ Days + (Days | Subject)
Data: sleepstudy
REML criterion at convergence: 1743.6
Scaled residuals:
Min 1Q Median 3Q Max
-3.9536 -0.4634 0.0231 0.4634 5.1793
Random effects:
Groups Name Variance Std.Dev. Corr
Subject (Intercept) 612.09 24.740
... 阅读全帖
d*****u
发帖数: 17243
6
这种研究在美国和中国大概都批不了
他们研究了欧洲常见的Y-chromosome halogroups与智商的统计关系
然后得出A组(包含R,N,I等)比B组(J,E,T,L)智商高
https://lesacreduprintemps19.files.wordpress.com/2012/05/haprinderm.pdf
5. Discussion
Based on our model, it appears that national cognitive
ability is confounded with the general development of
society. This is shown by the high correlations with HDI and
the observation that in three analyses HDI accounted for the
largest mean share of the variance in national cognitive
ability. The mean across two regres... 阅读全帖
f***c
发帖数: 301
7
来自主题: JobHunting版 - LinkedIn Data Scientist 面经 偏统计
第一次on-campus interview. 主要问了三个东西:
1.1个store 有10排队 每个队只有一个收银员,然后每个队伍有10个人在排队。另一个
store只有1排队,有10个收银员,有100人在排队,问你会选择哪个队伍来排,假设每个
收银员办理一个客户的时间和variance一样。 那两个Store的expected waiting time
是一样的
但是variance不一样 主要是分析mean-variance
2.问在java里“==”和.equals的区别 还有如果.equals要继承的话怎么写
3.数据分析。linkedin 现在卖一种license给其他公司(每个license可以看成一种产品
)。大概4或者6个公司的数据,每个data point里有三个variable,1.公司名字,2.# of
license purchased 3.unit license的价格. 先问能从数据里看出什么,
基本就是#of license purchased 和unit price正相关。但是这不能说明我们可以提高
unit price来增加# of license... 阅读全帖
A*******s
发帖数: 23
8
【 以下文字转载自 JobMarket 讨论区 】
发信人: AppleJobs ((can help refer)), 信区: JobMarket
标 题: [Apple Jobs]Power system, AC-DC, backlight driver
发信站: BBS 未名空间站 (Tue Feb 5 01:48:09 2013, 美东)
Send resume with position title to a****************[email protected]
Sr. AC-DC Designer
Sr. Power Systems Engineer
Sr DC-DC Power System Design Engineer
Sr. Backlight Driver Chip Definer
Sr. AC-DC Designer
Job Summary
Imagine what you could do here. At Apple, great ideas have a way of becoming
great products, services, and cus... 阅读全帖
t****d
发帖数: 89
9
来自主题: JobHunting版 - 请教版上大牛一道算法题

score总和 = (i+1)*(list[0]到list[i]数值的variance)+(j-i)*(list[i+1]到
list[j]数值的variance)+(n-j)*(list[j+1]到list[n]数值的variance)
r****r
发帖数: 66
10
来自主题: JobHunting版 - 请教一个概率计算的问题
假如 随机变量 X follows normal distribution (u, var), u is mean, var is
variance
y1=x^0.5 的mean 和 variance 是什么?
y2=x^0.25 的 mean 和 variance 是什么?
多谢各位指教, 不胜感激。
A*******s
发帖数: 23
11
Send resume with position title to a****************[email protected]
Sr. AC-DC Designer
Sr. Power Systems Engineer
Sr DC-DC Power System Design Engineer
Sr. Backlight Driver Chip Definer
Sr. AC-DC Designer
Job Summary
Imagine what you could do here. At Apple, great ideas have a way of becoming
great products, services, and customer experiences very quickly. Bring
passion and dedication to your job and there's no telling what you could
accomplish.
Key Qualifications
See below
Description
Job Descripti... 阅读全帖
o********n
发帖数: 1329
12
来自主题: JobMarket版 - job in Boston area
From: HR Communications
Sent: Friday, September 20, 2013 11:10 AM
To: +Billerica-All Users; +Hackensack-All Users; +Perceptive-East Windsor; +
Perceptive-Wayne; +Waltham-All Users
Subject: Featured Job Openings for Northeast [on behalf of Rose Barsumian]
Featured Job Openings for Northeast
Do you know the perfect candidate for a job at PAREXEL?
These Featured Jobs highlight open positions in your local area and there
may be others so please check the Employee Referral website for a full
listing ... 阅读全帖
c*******u
发帖数: 12899
13
来自主题: NextGeneration版 - [合集] 男孩智力真的只遗传母亲吗?
☆─────────────────────────────────────☆
karo (karo) 于 (Fri Mar 4 12:21:45 2011, 美东) 提到:
看到这样一篇文章,不知道有没有道理啊?如果是真的,我太伤心了,因为我们家都是
男宝宝,本来指望聪明的爸爸能把智商遗传给他们,现在我这个笨妈不是要害惨我的宝
宝们了。55
http://zhidao.baidu.com/question/5021375
智商的遗传基因ZT
满有意思的^^
据说决定智商的八对基因全部都是位于X 染色体上面,然后男生是XY,
X 是来自母亲,
Y 是来自父亲。
所以男生的智商全部都是来自母亲的遗传。
然后女生是X X,所以女生的智商是父亲跟母亲各有一半影响。
然后说,因为女生的智商是父亲母亲都有影响,所以会有中和的效应。
所以女生智商的分布会呈现自然分布( normaldistribution ),就是倒钟状,中间最
多,两边较少。
然后男生因为是完全只受一方影响,所以男生智商分布会呈现在偏向在两个极端。
也就是说,男生天才比较多,但是同时,蠢材之中也是男生特别多。
这个故... 阅读全帖
w****j
发帖数: 6262
14
嗯,可以换个算法,分别算先富的variance也很小,后富的variance也很小。
然后把两个方差加起来除以二,最后的variance还是很小。
所以中国的贫富差距很小。
n***r
发帖数: 2074
15
这个鸡巴区间看涨,看跌纯属扯淡,超过了看跌,低于就看涨,尼玛就两字,傻逼。
你算一个5日的MA均线,然后再算算股票的variance
5日均线加减variance乘以某个倍数K,这个K可以regression出来
这尼玛就是统计里最基本的区间估计,都能忽悠一帮子土鳖。
你们去算算吧,这种低级的预测都拿出来忽悠,实在是扯淡
这里几日的均线全屏自己规定,几倍的variance也可以自己规定,regression两个参数
就好了。
y***r
发帖数: 16594
16
来自主题: Stock版 - AAPL 这个股票长期FA
ACKMAN 的话,可以解释去年苹果股价的各种不如意。。。。
我早给你们概括了:“ It is dragged down by its own weigh, period.”
细节很复杂,包括什么growth to value呀,index fund buy when price up,index
fund sell when price down. The trend continue until value investors come up.
一句话就是: 老大不好当。
今年应该不一样了,把老大的位置让给GOOGLE就好了。。。
下面是转发。
Index Funds
Index funds and other passive managers have gained increasing market share
in recent years. Investing capital in funds and ETFs that track major market
indexes has recently been what one might call a “one way b... 阅读全帖
y***r
发帖数: 16594
17
来自主题: Stock版 - AAPL 这个股票长期FA
ACKMAN 的话,可以解释去年苹果股价的各种不如意。。。。
我早给你们概括了:“ It is dragged down by its own weigh, period.”
细节很复杂,包括什么growth to value呀,index fund buy when price up,index
fund sell when price down. The trend continue until value investors come up.
一句话就是: 老大不好当。
今年应该不一样了,把老大的位置让给GOOGLE就好了。。。
下面是转发。
Index Funds
Index funds and other passive managers have gained increasing market share
in recent years. Investing capital in funds and ETFs that track major market
indexes has recently been what one might call a “one way b... 阅读全帖
A*******s
发帖数: 23
18
【 以下文字转载自 JobHunting 讨论区 】
发信人: AppleJobs ((can help refer)), 信区: JobHunting
标 题: [Apple Jobs]Power system, AC-DC, backlight driver (转载)
发信站: BBS 未名空间站 (Tue Feb 5 01:48:38 2013, 美东)
发信人: AppleJobs ((can help refer)), 信区: JobMarket
标 题: [Apple Jobs]Power system, AC-DC, backlight driver
发信站: BBS 未名空间站 (Tue Feb 5 01:48:09 2013, 美东)
Send resume with position title to a****************[email protected]
Sr. AC-DC Designer
Sr. Power Systems Engineer
Sr DC-DC Power System Design Engineer
Sr. Backlight Dri... 阅读全帖
M*******A
发帖数: 14451
19
你这个是QTY Variance,Price Variance你没考虑到
我说的是Variance
A*****a
发帖数: 52743
20
这个不叫统计解释,这个勉强算概率解释
统计定量的是variance, 一个76.3%,84.3%是个平均数,后面缀着或大或小的variance
,
这个variance的存在,使得你所有的概率推论都没有任何实际意义。76.3%或与84.3%没
有任何统计上的显著不同。
w******m
发帖数: 13
21
来自主题: Football版 - powerball又没中
这个说法纯属扯淡。买lottery是用EV买variance,就像买保险是用EV卖variance一样
。人生不可重来,不是统计上的可重复试验,所以当variance足够大的时候,人们会做
出严重偏离EV的决定。这个道理我很久以前用很长的篇幅具体阐述过,懒得再多说。反
正看得懂的都懂,看不懂的说再多你也不明白。最后,买lottery的必须绝大多数是穷
人,因为边际效用递减原理。
c*****t
发帖数: 817
22
扔硬币已经确保了公平。其实楼主的本意是不喜欢随机性觉得TD结束比赛的variance太
大。
问题是没有能在尽快结束比赛的前提下减少variance的办法。
要想既公平又减少variance。干脆OT不要突然死亡再打半小时。这样输家没话说了。
但是赢家也只能躺着回家了。
f*****g
发帖数: 15860
23
来自主题: TexasHoldem版 - 这个玩法怎么样
well, this would kill the game, or at least the luck part.
if you two are in a coin flip situation, nobody will make $ because if we re
-deal 100, 1000 times, it'll exactly be a coin flip;
if you're way ahead, say, AA vs. KK, then KK will exacly lose 80% of time.
other situations are similar, you'll win/lose exactly as odds indicated.
bluffing will mostly be killed too.
run it twice is already doing a little of this to reduce variance. this
would kill variance forever. without variance, this gam... 阅读全帖
W********m
发帖数: 7793
24
来自主题: TexasHoldem版 - WHY AM I STUCK AT LLSNL? ZT (2+2)
Part I
It’s a question I ask myself all the time, why am I still at these levels? I feel like my thought process is definitely higher than 95% of live 200nl players, and probably a big portion of live 500nl players. I mean I can tell you what *enter famous players name here* is going to bet on HSP or WSOP. Ok, maybe that last one was a bit of a joke, but I know I used to feel like that. Seriously, though, why am I stuck in this place? Is it bankroll issues? Yes, somewhat. I do spend part of my b... 阅读全帖
c*****t
发帖数: 817
25
来自主题: TexasHoldem版 - Live 还是 online poker?
Live poker is much easier to win money if you can play enough volume and
have enough bankroll to sustain the variance. Otherwise online is better
since you can multitable and minimize your variance. Personally, I really
hate the variance at live poker. For example, you drove 2 hours to the
casino, and played 6 hours and lost $500 due to a bad beat, then drove back
in a bad mood at 1am late night. Feels really bad.
W********m
发帖数: 7793
26
来自主题: TexasHoldem版 - A sick fold on the flop
这不就是我在24楼说得吗?variance 大,因为对手的continuation range wider with
a drawy board. 你选择fold 来reduce variance, 我没问题。 see my post
"This is the bottom line:
1) more value with tptk on a drawy board.
2) more variance with tptk on a drawy board.
whether you want to continue to play your tptk with a drawy board depends on
whether you want to put your 400$ at risk to maximize your expected value.
Player chose "NO". 400 is too much to lose. I think this is totally fine"
不过你硬要说你的AQ is weaker facing a drawy board, ... 阅读全帖
c******q
发帖数: 456
27
来自主题: TexasHoldem版 - 考,还有这么打牌的
Variance in poker is the evil. Sometimes it makes you feel like Phil Ivey;
but most of time it makes you feel like idiot. A good player knows how to
adjust his game to conquer variance, but an excellent player knows how to
adjust his game to use variance as a weapon to conquer his opponents.
T********n
发帖数: 528
28
来自主题: TexasHoldem版 - 我也初试LHE并兼向潭老师请教

hold
不客气不客气,我非常乐意给意见。不一定永远是对的。越多人讨论一般越容易从理论
上找出答案。poker好玩的地方就是合逻辑又dynamic,而根据不同的情况不同的line会
net higher or lower EV。我只是比大部分同学们早个好几年开始打。我个人把poker
看成a game with an endless cycle of decision making, and better decision
leads to a higher score, and in this game, the score is money. [Eventually
it becomes also psychology and metagame, but that's far far down the road]
所以这其实就是你第二个问题的答案。Allow me to explain. 虽然你可以从数据和数
学来看你是不是beat a particular limit/game,可是现实上这不是那么容易答出来的
。 If you are playing online, w... 阅读全帖
q****8
发帖数: 3281
29
昨晚通宵,输了6K。纯粹TILT SPEW MONEY了2K,因为输钱而不在自己A GAME输了1-2K
,还有2-3K逃不掉。有几点希望大家发表些看法:
1. 输钱了肯定没赢钱的时候打的好,输大了就不在A GAME上了,这应该对于所有人都
这样吧,或许只是程度的差别?
2. 对有些人或许是你认为的FISH,在一个SESSION里你怎么打怎么输,不断充钱不断输
,然后越来越TILT,不断送钱。风暴兄你打了100万手牌了,应该碰到过吧,你能很及
时的退出吗?尽管你认为他是FISH,但今天他就是什么牌都HIT到。
3. 我赢钱了就打的比较小心,打法尽量降低VARIANCE。输大了就尽量建大POT,打法
VARIANCE越大越好。就算EV一样(其实应该不一样,输大了判断力会下降很多),
VARIANCE的区别会让你赢小钱的SESSION次数很多,输大钱的SESSION次数很少,但一次
就抵上好多次赢钱的SESSION.
y********n
发帖数: 2063
30
前提条件是,彼此之间信息全无
我的一点看法
1 以你所在的ps 25NL的整个player pool的信息为基础,做决定。比如大牛你前几万手
牌里,包括你involved或者是旁观的3bet pot的%,preflop是不是经常3bet 4bet,整
体的commit pot的hand strength,hand showdown的比例,整体的aggression程度(即
大多数时候,这个级别的game是soft或者是push hard)等等。哥觉得在相对soft的
game environment里,JJ在这里call 3bet oop,以pot control为基础play post flop
问题不大,换言之,marginal。而在非常pushy的game里,postflop会变得异常难打,
大多数情况下ended up猜对手有没有牌,大的variance不可避免。
----------Yes, you are right.
而在非常pushy的game里,postflop会变得异常难打,
大多数情况下ended up猜对手有没有牌,大的variance不可避免。
2 从game... 阅读全帖
d*****0
发帖数: 1500
31
来自主题: TexasHoldem版 - what would u do in 哥的 spot here

later
This is actually a good point.
先申明,我不是在defend我的move,否则下面的话,就没有意义了。
珍珠哥觉得,既然对对手有很大的优势,就应该放弃已经invest的50个bb,来避免这个
非常非常大的variance。因为后面有的是机会,以技术优势,一点一点把对手的stack
赢过来。
我有两条理由来表示不同的看法。
1 这手牌确实存在巨大的variance,但是我的理解是,对手很有可能以非常松的range
,在我的平call之后,allin,也就是说,实际是我自己创造出这样一个机会,以领先
对手range的情况下get in。当然,换来的是huge risk/huge reward。换言之,如果不
想以这种方式gamble的话,我就直接选择limp/fold了。
2 以技术优势,一点一点用更稳妥的方式(或者说low risk/small reward)攫取对手
的筹码。这点没错,但是跟这个spot实际上是两个独立事件,前提是我的bankroll足够
take这个huge variance。
另外,我还有两个小原因,一是建立自己的toug... 阅读全帖
p****0
发帖数: 611
32
来自主题: TexasHoldem版 - what would u do in 哥的 spot here
Nothing personal. First of all, I don't think your call had anything wrong
with it. It's totally a right/standard move, if you are ready to take big
variance. Event with any two, you can make that call, because the image you
described about the BTN is about any two style and gambler. His shove
against you on this spot didn't give you any information about what hand he
has. Most likey, he had nothing. So you can always play coin flip with him.
Let me ask in a different way. If villain has 27o... 阅读全帖
d*****0
发帖数: 1500
33
来自主题: TexasHoldem版 - To gamble or not to gamble
我觉得老版主get the point,然后ak哥扯到别的地方去了。
如果ak哥真心想protect the capital and avoid variance,limp/flat with suited
small ace就失去了它的point,从长远来看,就是慢慢地一次一次地donate你的这种情
况下的6个bb。要知道,捐15次,也就相当于一个buyin了,且这十五次里,有多少次能
够看到类似的dream flop?很多时候,一点不击中还好,要是击中一个a或者7什么的,
还得继续兜进去chips,真正的在那么多case里,能够成为winner of the cooler,把
对手一锅端,赢下200bb的概率,实在太小。反倒是,人真正能跟ak哥commit 200bb的
hands,往往领先于flush以外的A7 made hand。
换句话说,玩非premium hands本身就代表了追求implied odds和不可避免的variance
。我想ak哥打牌,主要还是喜欢牌局上的小搏弈,偶尔赢点小钱,心里无限
comfortable。如果是这样的话,真的没必要,去玩那些,输了会... 阅读全帖
d*****0
发帖数: 1500
34
来自主题: TexasHoldem版 - 晒一下负能量


这个哥有不同看法,决定variance的因素有二,一是volume,二是技术差距,即edge。
相同volume的情况下,技术差距越大,variance越小。举个极端的例子,叫一个新手玩
家拿50个buyin去跟ivey headsup一天,你说ivey被对手wipe out的几率有多大?
反过来讲,如果edge不是那么大,那就变成了老湿所说的,因为volume不够,打一辈子
都没办法搞清楚自己的水平到底是高还是低。
而实际情况是,同一级别,大多数玩家水平都差不了多少,偶尔有新手加入,也是早早
的被分食,或者再也不打牌,或者去低级别的game从头开始学。
所以,真的想要正儿八紧地没有variance的赢钱的话,反而可以去玩最低级别的桌子,
开他18桌,只玩qq+/ak,不中就check/fold,中了就pot pot pot。无脑收钱!
比起高级别的桌子,打了几年发现持平或者还输者,那是好上了一万倍。
W********m
发帖数: 7793
35
来自主题: TexasHoldem版 - 晒一下负能量
My 2 c
Ev 只是你在河牌出来前, all in 时的ev 线离true money earning 的距离。这只是
你run good or bad 的一部分, 也是表现poker variance 的一部分而已。其它的
factor, 比如你拿到多少次AA, AA到河牌hold 了多少, above or below expected
value, 都会让你的ev 线和真实win rate 有偏差, 也是ev 线上看不出来的。poker
variance 的factor 太多了, 这也是为十么poker variance 大的原因。

★ 发自iPhone App: ChineseWeb 8.6
p******a
发帖数: 975
36
来自主题: TexasHoldem版 - One hand
这个道理说起来容易但是做起来很难。比如我虽然知道这个strategy绝对是正确的,但
是当我在bovada上开始run bad的时候(比如现在),就会不由自主的减小bet size,
牺牲win rate来减小variance。我还仅仅是在玩NL200,如果在更高的级别,面对更强
的对手势必会有更大的variance,而且涉及的钱也更多,就更难了。
所以你看isildur一个波动几百万,其实真心不大。他玩比如NL40k的级别,一百万不过
是30个buy-in。考虑到他疯狗一样的打法,而且还多是单挑,其实他的variance比
scripps还小
W********m
发帖数: 7793
37
来自主题: TexasHoldem版 - pokerstar 到底有没有问题
扑克的variance 是很大,但也没有你想象的那么大。
至少在NL200 以下你完全可以玩比较稳的打法,而没有很大的浮动。 我的一百万手牌
里 (NL10- NL200 都有),最大的down swing 也就10-15 个 buy in而已, 好像也就
只有三次。当然这和打法有很大的关系。
如果你有经常在一个级别经常有20-30 buy in 或更多的down swing, 你可能要深思一
下,
1)你到底是不是这个游戏的真赢家, variance 大的时候,你对自己的真实赢率的估
计会有很大出入。
2)你的玩法的variance 和ev 的比例可能并不值得你玩这个游戏(破产可能和赢率比
)。
w***w
发帖数: 6301
38
来自主题: TexasHoldem版 - 请教SNG怎么提高
我早期就是只打SNG。
现在只打cash。
我一点不觉得SNG 的variance 小。
作为一个game品种来说,SNG不利于体现技术差距,不利于体现edge的作用。
variance 最小的其实是cash HU。因为这种game最利于发挥edge。
但是对新手来说,他可能打SNG的起伏比cash game小。那是因为新手往往都会去打SNG
。换句话说,他在SNG遇到的对手水平比cash game遇到的对手水平要差。所以他能有比
较大的edge。但是这种variance小是因为对手差给他的edge,而不是game品种给他的
edge。所以你在高级别的SNG中,就会比高级别的cash更难取得edge。
说cash game是金饭碗,还有一个意思是说cash game你打好了,到处都能玩。
home game,赌场,online随便哪一个小site,你都不会找不到台子。
你打SNG,要立刻有台子上,online现在可能只有PS了。
在WPN和chico这样比较大的US facing sites,你要玩SNG都要等很长时间。
在EPN一台SNG凑足人要几个小时。其他sites就更不用说。
... 阅读全帖
w***w
发帖数: 6301
39
来自主题: TexasHoldem版 - 请教SNG怎么提高
如果把打破bankroll风险的概念从variance中除去,那么我们讨论poker的variance还
有什么意义呢?大家为什么又要避免variance呢?
这两者根本就是一体的。
q****8
发帖数: 3281
40
来自主题: TexasHoldem版 - 请教SNG怎么提高
避免VARIANCE很重要的是心理方面的原因。VARIANCE大了,总有越打越糟的时候,碰到
大的DOWNSWING,虽然此时自身的原因已经超出了客观随机波动的原因,但大的
VARIANCE是诱因。
v*******e
发帖数: 11604
41

你懂啥叫 variance?variance大的意思是极端情况出现机会多。所以优选出来的聪明
人可以很聪明。
另外讨论一下前面有人说的的纯种交配,那是要保留某些特别的feature,比如斑点狗
、折耳猫之类。如果你找陶哲轩来搞纯种交配保留聪明基因那是很好。如果用普通人,
那么还是后代variance大才容易优选出极端情况的人。
g**a
发帖数: 2129
42
来自主题: Joke版 - 简单的数据处理问题
9 samples out of 500 or 9 samples out of population?
If it is the second one, assuming your 500 samples are iid and random, you
should be able to use m and s to estimate sample mean and variance. The dist
for 9 samples mean should also follow normal dist with mean and variance
calculated from estimated population mean and variance.

The
accurate
what
A*********u
发帖数: 8976
43
来自主题: PKU版 - here u go
Whereas one-way analysis of variance (ANOVA) tests measure significant
effects of one factor only, two-way analysis of variance (ANOVA) tests (also
called two-factor analysis of variance) measure the effects of two factors
simultaneously. For example, an experiment might be defined by two
parameters, such as treatment and time point. One-way ANOVA tests would be
able to assess only the treatment effect or the time effect. Two-way ANOVA
on the other hand would not only be able to assess both time
k**********g
发帖数: 989
44
请问楼主,公式(2)是要让任何正整数moment (q) 成立,还是只需要其中几个(q)值成
立?(eg. q=1 and q=2 only)
楼主测试时,用了那几个(q)值?能否列出来让买买提的菜鸟看看。
我没数学背景,所以只知道(q=2)的特例。如果楼主真要知道(q!=2)的话最好到
Computation问。
(q=2)可以用两种思路理解。 第一种思路是 intra-class variance vs inter-class
variance. Partition 後, Variance 会符合某恒等式。http://en.wikipedia.org/wiki/Otsu's_method
第二种思路是 2D fourier transform。把图像缩小2X,等於把fourier domain的3/4的
frequency bin 去掉,剩下1/4的frequency bin。
n******7
发帖数: 12463
45
最近看java,看了快200页还没扯到String,烦了
因为我的工作设计很多文件parsing的事情,就直接开写来学习
结果卡在map上了,求给个示范
经常要处理文件,比如input.txt:
Group,Sample,Gene,Variance
Patient,1,A,V22E
Patient,1,A,D54N
Patient,1,B,K17R
Control,2,C,A97G
Patient,3,A,V22E
Patient,3,C,I31V
Patient,3,D,G85A
Control,4,C,A97G
用python处理是这样的:
data = {}
with open('input.txt') as f:
f.next() #skip the title line
for line in f:
(group,sample,gene,variance) = line.rstrip().split(',')
if group not in data:
data[group] = {}
i... 阅读全帖
p*****2
发帖数: 21240
46
来自主题: Programming版 - Scala又被鄙视了
Scala's disjoint part has polymorphic sub-typing and implicit dictionaries.
Haskell's disjoint part has first-class type classes, type-level functions,
full type inference, and many other neat little tweaks. The intersection is
a higher-kinded type system with first-class functions.
Scala's part of the intersection, for a list of reasons that include JVM
type erasure, issues combining higher-kinded types with polymorphic sub-
typing, and certain implementation choices, is partially "broken" in t... 阅读全帖
p*****2
发帖数: 21240
47
来自主题: Programming版 - 我觉得在scala上浪费时间没意思
Scala's disjoint part has polymorphic sub-typing and implicit dictionaries.
Haskell's disjoint part has first-class type classes, type-level functions,
full type inference, and many other neat little tweaks. The intersection is
a higher-kinded type system with first-class functions.
Scala's part of the intersection, for a list of reasons that include JVM
type erasure, issues combining higher-kinded types with polymorphic sub-
typing, and certain implementation choices, is partially "broken" in t... 阅读全帖
f********x
发帖数: 99
48
The world beyond batch: Streaming 101: A high-level tour of modern data-
processing concept
http://radar.oreilly.com/2015/08/the-world-beyond-batch-streami
by Tyler Akidau August 5, 2015
Editor’s note: This is the first post in a two-part series about the
evolution of data processing, with a focus on streaming systems, unbounded
data sets, and the future of big data.
Streaming data processing is a big deal in big data these days, and for good
reasons. Amongst them:
Businesses crave ever more tim... 阅读全帖
f***0
发帖数: 102
49
$2B public mfg company operational finance 方向( cost accounting, annual
operation plan, actual vs. budget variance analysis),belongs to "
procurement to pay" cycle, has nothing to do with "revenue to cash" cycle.
level 5(最高技术level非MANAGER), requires extensive knowledge of operation
process and its reflection in accounting for each operation/mfg transitions.
Require knowledge of Oracle functional modules, query writing and simply
programming skill required. maintain integrity of prepaid/consigned... 阅读全帖
f***0
发帖数: 102
50
$2B public mfg company operational finance 方向( cost accounting, annual
operation plan, actual vs. budget variance analysis),belongs to "
procurement to pay" cycle, has nothing to do with "revenue to cash" cycle.
level 5(最高技术level非MANAGER), requires extensive knowledge of operation
process and its reflection in accounting for each operation/mfg transitions.
Require knowledge of Oracle functional modules, query writing and simply
programming skill required. maintain integrity of prepaid/consigned... 阅读全帖
首页 上页 1 2 3 4 5 6 7 8 9 10 下页 末页 (共10页)