w*********s
发帖数: 2136 | 1 Jeffry John Aufderheide
vactruth.com
05/25/2010
http://vactruth.com/2010/05/25/vaccines-rockefeller-social-control/
PART I.
History books proclaim with absolute certainty that the pinnacle of public
health is defined by the polio vaccine program. Disease finally conquered by
science. The polio effort was a benchmark for the public to mentally accept
the concept “shots prevent disease.” However, hidden to most Americans
was an elaborate Public Relations scheme being carefully applied by
Rockefell... 阅读全帖 |
|
n********k
发帖数: 2818 | 2 Concerning Liu's case, Essentially asses dictate brains...He is certainly
not someone every PI should strive to be...But frankly I don't see the
rationales some mad-dog-like IDs are going after him either...If one reads
through his stories: he is likely a very nice/caring and upright but tough
individual with integrity while he might have a short fuse or some
personality problems and could use some interpersonal skills in his
management...And his lab could be a great place for those who have rea... 阅读全帖 |
|
C****1
发帖数: 89 | 3 I personally agree with your firs part but disagree with the second part.
When I read this "I personally don't think any real BT or even not so bt PI
would allow his/her students to challenge his/her authorities and then get
away with it that easily...I don't think the student would dare to do it in
such a childish way if he/she predicates there would be a bad consequence
for him...", my understanding is like that student should deserve to be "
killed" instead of simply being kicked out of the p... 阅读全帖 |
|
n********k
发帖数: 2818 | 4 Has he win a Nobel yet? he knows everything and predicates everything, and
he has a journal too...kind of sad for him...it is really sad, and he could
have been a fine scientist... |
|
L********d
发帖数: 3820 | 5 bioinfo不就是一直在用data mining的各种技术么？
算法再怎么改都比不过你的数据量多有用，high dimension curse主要是data数量不够
要想build一个好predictor，有足够精确的training sample才是王道
bayesian statistics和classic方法的区别就是bayesian combine了prior knowledge和
observations，什么叫综合了很多databases?
bayesian methods早就已经广泛并大量地应用在bioinfo中了
随着一个样品的各种类型的测序结果越多，high dimension curse 应该更大了。我觉
得即使是sparse 的各种算法也只能改进一小部分。看理论分析和simulation，个人认
为算法的改进对最终的predic |
|
n*********m
发帖数: 38 | 6 I think tonight Noble prize will be award to one of the following research
topic/scientist:
left-right asymmetry theory/Hiroshi Hamada etc,
signal transduction/Lewis C Cantley, Tony Pawson (?) etc,
or autophagy/ one from Japanese,
insertional transposon mutagenesis/AC Spradling & GM Rubin. |
|
l*****n
发帖数: 214 | 7 Tony Pawson 都去世了。。。。。。。。 |
|
n*********m
发帖数: 38 | 8 That's why there is a (?). |
|
o*****e
发帖数: 311 | 9 I bet on Spradling and Rubin! |
|
s*****j
发帖数: 6435 | 10 我看应该开发一个 iq predicator
看看做什么研究的对智商伤害最大. |
|
p*******r
发帖数: 4048 | 11 Yes, this hypercompetition is what made me decide to leave this country.
I dreamed with another friend from MITBBS to build a company predicated on
standardized data production and bioinformatics and a university lab free
from such strains.
I am glad that he has succeeded in quite a big way and I had some partial
success.:)
the |
|
s******s
发帖数: 13035 | 12 你PC1有12%，不知道你为啥认为低了，和啥比的？
你这个PC的variance explained降的很快，我不错metagenomics，不过从统计上讲，这
说明前两个PC很可靠啊。
你用PC做predication或者cluster，我知道很多人这样做，结果出来也好解释，纯生物
的喜欢，不过这玩意儿其实就是看看啊，真要做肯定直接放model啊。PCA这玩意儿，和
distance function关系太大，不一样的data transformation, 不一样的distance
measurement结果可能完全不同，也就一fast & dirty的方法，真不适合深究。 |
|
H*****l
发帖数: 702 | 13 CdTe and CIGS sucks
solar people make many many many predication of god damn promising future
did any of them happen indeed?
NREL almost cut off all CdTe research except their own group and USF
CIGS...if they are still growing CIGS on glass not flexible substrate like
plastic, there is no better than a-Si tandem
And till now, did Nanosolar make any production out?
the only promising technology here is silicon if considering large area/
affordable cost
CIGS |
|
H*****l
发帖数: 702 | 14 CdTe and CIGS sucks
solar people make many many many predication of god damn promising future
did any of them happen indeed?
NREL almost cut off all CdTe research except their own group and USF
CIGS...if they are still growing CIGS on glass not flexible substrate like
plastic, there is no better than a-Si tandem
And till now, did Nanosolar make any production out?
the only promising technology here is silicon if considering large area/
affordable cost
CIGS |
|
h*****k
发帖数: 744 | 15 那要看你对"好文章"的标准是什么了. "好文章"都需要大量的时间和精力的,"好出结果
"的东西,并不一定容易出"好结果".俺上面说环境化学出数据多,也不大准确,要看具体
方向了.化学方面做吸附的,多半都会加个model上去. 至于环境微生物方面,俺看到有
model的paper一般都直接跳过.俺觉得很多看着不错的model都是在相当严格的条件下,
人为地控制了很多参数,可实际上的环境都复杂的多.把model用做rough predication还
行,比如废水处理厂的运行.
另外,"量"和"质"很难达到同步的,求"量"的时候,很难保证文章的"质". 不过话又说回
来,谁的paper里不都是有些好的,也有些属于灌水性质的,也算正常. |
|
m******t
发帖数: 273 | 16 【 以下文字转载自 Statistics 讨论区 】
发信人: myregmit (myregmit), 信区: Statistics
标 题: Predict values of vectors by other vectors generated by black box
functions
发信站: BBS 未名空间站 (Sun Mar 2 11:38:56 2014, 美东)
Hi,
I need to solve a problem about predicting values of some numerical vectors
by using other numerical vectors with all these vectors in the same vector
set, which is generated by one or more black box functions.
Given a vector space:
P =[S_1, S_2, …, S_T | Sk is a vector of q numbers, k = 1, ..., T]... 阅读全帖 |
|
m*******1
发帖数: 328 | 17 分数不是everything, 但是也不是nothing. 毕竟programs avidly screen candidates
based on the scores. Scores have predicative values with regard to your
final board pass rate and your performance during residency in some
specialites.
我虽然不是PD, 一些PD认为对于IMG:
Scores reflect IQ+knowledge (trainable or not).
IVs reflect EQ+interpersonal skills(sociable or not).
Research reflect your dedication(devoted or not).
Clinical experience reflects your job related experience (hands on or not). |
|
w*********s
发帖数: 2136 | 18 Jeffry John Aufderheide
vactruth.com
05/25/2010
http://vactruth.com/2010/05/25/vaccines-rockefeller-social-control/
PART I.
History books proclaim with absolute certainty that the pinnacle of public
health is defined by the polio vaccine program. Disease finally conquered by
science. The polio effort was a benchmark for the public to mentally accept
the concept “shots prevent disease.” However, hidden to most Americans
was an elaborate Public Relations scheme being carefully applied by
Rockefell... 阅读全帖 |
|
c******8
发帖数: 13 | 19 Well, if this is the stock market that I am trying to predicate, I wouldn't
be
using this. I am using this simply b/c I don't really care who wins.
the |
|
t*******y
发帖数: 11968 | 20 【 以下文字转载自 NewYork 讨论区 】
发信人: talkdirty (talkdirty), 信区: NewYork
标 题: Bloomberg terminal 问题
发信站: BBS 未名空间站 (Wed Feb 11 21:16:34 2009)
运行过predicated cash flow on an uploaded portfolio以后, 如何把report export to excel?
谢谢. 我问了BB tech,他们居然说没这功能,我不敢相信... |
|
p********0
发帖数: 186 | 21 I heard the contrary to "Historical volatility is better than implied". A
prominent prof mentioned in his class said he tried GARCH/ARCH/ARMA
everything, but turn out to use the implied volatility which is more
accurate than the historical one.
Assume Implied volatioity is 100 percent accurate, than by minimize the
implied volatility for a portfolio means the future volatility is minized.
In case error is big(predication power is weak) then minimize doesnot mean
anything useful. |
|
m******t
发帖数: 273 | 22 【 以下文字转载自 Statistics 讨论区 】
发信人: myregmit (myregmit), 信区: Statistics
标 题: Predict values of vectors by other vectors generated by black box
functions
发信站: BBS 未名空间站 (Sun Mar 2 11:38:56 2014, 美东)
Hi,
I need to solve a problem about predicting values of some numerical vectors
by using other numerical vectors with all these vectors in the same vector
set, which is generated by one or more black box functions.
Given a vector space:
P =[S_1, S_2, …, S_T | Sk is a vector of q numbers, k = 1, ..., T]... 阅读全帖 |
|
u***************r
发帖数: 11227 | 23 【 在 blaze (blaze) 的大作中提到: 】
I don't see how this article explains any of Godel's
Incompleteness Theorem (GIT for short). GIT has
nothing to do with cardinality of N, Q and R.
GIT says that there exists true but unprovable "first-order
logic" (or predicate logic) sentences in the structure
(N,0,S,+,*), where N is set of natural numbers, 0 is zero,
S is successor function, + and * are addition and multiplication.
This unprovable sentence just says something like "I am unprovable."
The contributi |
|
m******t
发帖数: 273 | 24 【 以下文字转载自 Statistics 讨论区 】
发信人: myregmit (myregmit), 信区: Statistics
标 题: Predict values of vectors by other vectors generated by black box
functions
发信站: BBS 未名空间站 (Sun Mar 2 11:38:56 2014, 美东)
Hi,
I need to solve a problem about predicting values of some numerical vectors
by using other numerical vectors with all these vectors in the same vector
set, which is generated by one or more black box functions.
Given a vector space:
P =[S_1, S_2, …, S_T | Sk is a vector of q numbers, k = 1, ..., T]... 阅读全帖 |
|
m******t
发帖数: 273 | 25 【 以下文字转载自 Statistics 讨论区 】
发信人: myregmit (myregmit), 信区: Statistics
标 题: Predict values of vectors by other vectors generated by black box
functions
发信站: BBS 未名空间站 (Sun Mar 2 11:38:56 2014, 美东)
Hi,
I need to solve a problem about predicting values of some numerical vectors
by using other numerical vectors with all these vectors in the same vector
set, which is generated by one or more black box functions.
Given a vector space:
P =[S_1, S_2, …, S_T | Sk is a vector of q numbers, k = 1, ..., T]... 阅读全帖 |
|
|
o******6
发帖数: 538 | 27 ☆─────────────────────────────────────☆
restart (testit) 于 (Sun Mar 15 17:35:19 2009) 提到:
请问：
同一组用户，每个用户都先后用两种工具(A,B)完成同一个任务。
其中一半用户先用A完成任务, 再用B完成任务；
另外一半用户先用B完成任务, 再用A完成任务.
我的目的是比较那种工具更方便使用一些.
请问在做t-test时, 应当选用 paired samples t-test 还是
independent samples t-test.
非常感谢!
☆─────────────────────────────────────☆
lightyear11 (lightyear) 于 (Sun Mar 15 18:01:20 2009) 提到:
The respond variable is how well the target is completed(maybe time?).
I think here you have 3 predic |
|
a********a
发帖数: 346 | 28 In proc lifetest, how to overlay the graph of KM from real observed survival
time and KM from predicated survival time?
Thanks |
|
a********a
发帖数: 346 | 29 I attached my sas code and graph in the attached file. I have some question
to make the graph a little bit nicer. Do you know how to do it?
Question:
1 how to make the predication line and confidence interval line a little bit
thicker?
2. How to make the value in Y-axis and X-axis only keep 1 decimal place(i.e.
remove the extra 0) and make the value shows a little bit bigger?
3. How to make the horizontal line 1.0 a little bit thicker?
goptions reset=all;
axis2 minor=none label=(font=swissb h=2 |
|
W**********E
发帖数: 242 | 30 本想按下列步骤做一个关于PROPENSITY SCORE的SIMULATION，但有问题，不知道哪位高
手能指点一下？
１） LP=B０+B１＊X１+B２＊X２+。。。。。+B１０＊X１０+
RAND（＂NOMRAL＂，０，１）＊０。３；
２） P=EXP（LP）／（１+EXP（LP））；
３） GENERATE ０／１ DUMMY TREATMENT GRP USING BERNOULLI（P）
４） GENERATE OUTCOME
Y=C０+C１＊X１+C２＊X２+C３＊X３+。。。。+BETA＊GROUP；
然后根据上面SIMULATED的数据，做分析
５）先用PROC LOGISTIC；
MODEL TRT=X１ X２ X３ 。。。。X１０；
输出PREDICATED PROBABILITY（PROPENSITY SCORE）。
６） 然后做模型；
模型１）
Y=GROUP PROPENSITY；
模型２）
Y=GROUP X１ X２ X３。。。。X１０；
两个模型的GROUP的系数差太多，模型１）差的很大，和PROPENSITY SCORE的作用想反
。我感 |
|
d*******o
发帖数: 493 | 31 (1)Logistic regression: I bet the dataset is heterogeneous(categorical +
numerous) and many values are missing. Have to do a lot of variable
aggregation, transformation and imputation. Amazingly still many ppl in
direct marketing are using it. Simply don't know why.
(2)Decision tree: CART/CHAID/C5.0/boosting and bagging
(3)SVM: way too slow. To run 50k+50K would kill most machines.
(4)Neural network: hard to explain.
Conclusion: the final predicative model would be an assemble of multiple
decis |
|
d*******o
发帖数: 493 | 32 保险公司的predicative modeler都是做CRM的，换到别的行业就叫database marketing
analyst. 需要跟account manager合作，属于marketing department,不属于actuary
department，公司一般sponsor考试。 |
|
F****n
发帖数: 3271 | 33 Obviously, you don't understand what I say.
SVM's non-linearity != Polynomials
SVM is superior because it handles more complex non-linearity with greater
generalizability.
If you only look at predication accuracy SVM will be "beaten" by a lot of
methods. But that's nonsense. |
|
c***i
发帖数: 15 | 34 pls send email to xwang2 AT ebay DOT com if you are interested.
Search Metrics, Staff Data Mining Analyst
Job Responsibilities
eBay Search Metrics is building world-class engineering teams in Shanghai,
focusing on search related systems and services such as user behavior/intent
analytics and crowd sourcing judgments. The team’s work will directly
impact
eBay revenue by improving both search engine performance and user
experience on eBay.
We are seeking a data mining analyst to provide applied re... 阅读全帖 |
|
m******t
发帖数: 273 | 35 Hi,
I need to solve a problem about predicting values of some numerical vectors
by using other numerical vectors with all these vectors in the same vector
set, which is generated by one or more black box functions.
Given a vector space:
P =[S_1, S_2, …, S_T | Sk is a vector of q numbers, k = 1, ..., T]
Find a sub-group of vectors g = {S_d | d belongs to 1, …, T } and a
function h( S_d )
Such that
Difference between the value of h(S_g) and S_r is minimized
set g + set r = set P
... 阅读全帖 |
|
k****n
发帖数: 165 | 36 Got it. Thanks! For your question,
1. For one predictor variable anova, variance = sum(y_ij-y_bar)^2
if predictor variable X is categorical, it makes sense to rewrite it as sum(
y_ij-y_i.+y_i.-y_bar )^2, where y_i. is the average for group level i; y_
bar is the grand average for the total observations; y_ij-y_i. is the within
group variability
while y_i.-y_bar is the between group variability (group is defined by X).
If X is continuous, it defines infinite many of such groups. Such technique
to... 阅读全帖 |
|
E**********e
发帖数: 1736 | 37 就是R里面的boxplots（）啊。 对每个predicator做个boxplot， 然后删掉outliers，
可以用1.5~3 std， 取决于你的samples size。 你需要写个小程序来删掉这些
outliers。 另一方法是regression啦，如果outlier是也就一两个的话，看看统计书。 |
|
m****s
发帖数: 18160 | 38 ☆─────────────────────────────────────☆
whiteclouds (/ 参考消息 /) 于 (Tue Mar 22 18:47:29 2011, 美东) 提到:
投诉人： Whiteclouds
投诉对象：我爱宝宝版主mitbbx及版副catdoudou
投诉标题：投诉我爱宝宝版主mitbbx及版副catdoudou大搞新闻网禁误导广大家长
投诉目标：罢免
投诉理由：
1. 事关每个人身心健康。你们凭什么封禁医药界的负面消息和非主流消息？
2. 关于疫苗利弊的讨论早已进入主流媒体了。你凭什么不让大家看到？
3. 药品（包括疫苗）近年屡有recall. 你凭什么禁止有关信息来误导家里有小宝宝的父母去盲目迷信药品质量？
4. 美加庸医遍地都是，医疗事故屡见不鲜。你凭什么制造版面气氛让家长对处方和诊断不过脑子的去盲信？
5. 凡事要听多方面意见，有点科学精神，你们动辄删贴封人缺乏领导版面的基本素质
6. 为人父母要首先自己要学会尊重不同意见，不然以后怎么教育好孩子？
7. 综上所述，建议罢免！
发信人: whiteclouds (/ 参考消息... 阅读全帖 |
|
l****t
发帖数: 36289 | 39 look仔细喽：
1.真凶是subject
现形是V predicate。
means：真凶现出原形。
2“真凶现行”不是完整语句，不能完整表达意思，即使makes sense的时候，也means
，真凶正在committing whatever bad deed.
你说的“真凶现行（犯）”语法不成立，语义不make sense。
看他回贴前的语境，是表达1的意思，所以是“现形"不是“现行”。监定完毕 |
|
s**********8
发帖数: 25265 | 40 http://www.qrasupport.com/FDA_MED_DEVICE.html
Overview: FDA Regulation of Medical Devices
The following information is provided as general guidance to the Food and
Drug Administration (FDA) regulation of medical devices. The Center for
Devices and Radiological Health (CDRH) is the division of FDA responsible
for medical device regulation. CDRH maintains a very informative web site.
The CDRH web site is located at: http://www.fda.gov/cdrh/devadvice/default.htm.
For support in working with the FDA... 阅读全帖 |
|
b********f
发帖数: 20 | 41 这也许能部分解释医疗总体费用，但是解释不了为啥单项检查或者procedure的收费那
末高。前面有人抱怨说虽然账单12万，可是最后保险只付了6万。如果保险只付6万就能
cover所有医护人员的薪水，成本，医院还有很大的利润，为啥医院出bill的时候狮子
大开口要12万？病人在接到账单之前大多不知道相关的费用，相关的收费信息并没有在
upfront 就提供。
医生也应该要通过市场的优胜劣汰，还有医疗行业的收费也要通过市场来调控价格，现在是通过人为造成shortage来达到价格控制。就像antitrust 不是说所有的monopoly都非法，而是通过predicate act来取得的monopoly是非法的。如果是一个自由的市场，最后病人愿意支付那末高的价格，那才是合理的价格。
但是现在由于医生的数目受到限制，造成医生的短缺，连医生都不够，何谈优胜劣汰,何谈自由的市场？ |
|
f**d
发帖数: 768 | 42 这是一本计算神经科学的优秀著作，全文拷贝这里（图和公式缺），有兴趣的同学可以
阅读
如需要，我可以分享PDF文件（－－仅供个人学习，无商业用途）
From Computer to Brain
William W. Lytton
From Computer to Brain
Foundations of Computational Neuroscience
Springer
William W. Lytton, M.D.
Associate Professor, State University of New York, Downstato, Brooklyn, NY
Visiting Associate Professor, University of Wisconsin, Madison
Visiting Associate Professor, Polytechnic University, Brooklyn, NY
Staff Neurologist., Kings County Hospital, Brooklyn, NY
In From Computer to Brain: ... 阅读全帖 |
|
h*****n
发帖数: 90 | 43 我对车本身其实基本不懂，猜他是指86比较容易oversteer，并且是属于predicable和
easily control/handle的那种oversteer，fun oversteer。从他的个人喜好来说，他
更享受86在赛道上漂移的控制感，所以我感觉他并不是很喜欢那种过弯出弯车子强grip
带来的快感。 |
|
z****m
发帖数: 92 | 44 read the first of the GS 10 predications for 2010 |
|
S******t
发帖数: 8388 | 45 what if it bounce up and then drop? |
|
|
|
|
|
|
