y*********g 发帖数: 239 | 1 It is mostly Logistic regression in risk modeling. Occasionally you will put
a few linear regressions in the process, but not often. I think the most
important aspect in risk modeling is to understand what you are doing –
meaning knowing your data and how to interpret the results. Credit risk is a
highly regulated industry with relatively less innovative technologies. A
lot of commercial software packages are now available that anyone can press
a button to build a logistic regression model along... 阅读全帖 |
|
y*********g 发帖数: 239 | 2 It is mostly Logistic regression in risk modeling. Occasionally you will put
a few linear regressions in the process, but not often. I think the most
important aspect in risk modeling is to understand what you are doing –
meaning knowing your data and how to interpret the results. Credit risk is a
highly regulated industry with relatively less innovative technologies. A
lot of commercial software packages are now available that anyone can press
a button to build a logistic regression model along... 阅读全帖 |
|
v*******a 发帖数: 1193 | 3 我最近对这块挺感兴趣的,想跳槽去这一块做,但是没有什么experience,希望得到有
经验的大侠的帮助。
有经验的朋友能不能帮我回答一下下面这些问题:
1) 做marketing mix model 是不是要用很多time series ? 用time series 的目的
是不是用来算 就算没有marketing campaign的baseline?
2)我上学的时候只学过longitudinal data analysis(都是cross sectional) 没学
过time series ,time series 都是自学的。 请问marketing mix model 里主要用
time series 的那些model?
3)大家觉得做marketing mix model 的前途如何?
本人目前在一个IT公司做做model,但是做的model 都是很industry specific的,用的
也就是linear regression logistic regression cluster 等。个人感觉这些技术也
不是很难,积累的经验大部分也只适用与目前所在的indust... 阅读全帖 |
|
A*******s 发帖数: 3942 | 4 我的两分钱--这是个面试必问的问题,我觉得答好不容易,答得全面了面试官还会容易
confused...
text说的都没错,但是需要了解为什么这么说的前提,以及实际面对的data和model的
性质。比如说“with multicollinearity, model's estimates are still unbiased
but have higher variance.” 这话自然没错,但是unbiased estimates的前提是你的
model没有mis-specification,或者mis-specification没有大的影响。如果real
model有10个variable但你只包括了9个,有bias。如果real model是nonlinear的但你
搞了个linear的,有bias。而且real model很可能是由一堆latent variables'
nonlinear functions组成的,在这种情况下,即使你把所有能观察到的variable都加
进去,也不能把所有的latent variables和他们的functional forms都包括到,m... 阅读全帖 |
|
A*******s 发帖数: 3942 | 5 model validation应该会涵盖银行的所有model
作为一个在risk和aml呆过的model developer来说,
除了你说的logistic regression之外,
我搞过的东西有--
model and simulate panel data with temporal and spatial correlation;
competing risk Cox model;
various forecasting models with exogenous variable;
GLM & Double GLM;
Copula;
retrospective case-control matching;
likelihood based missing data analysis;
text clustering and classification
所以说银行的model没你想的那么单调,
不过这完全取决于老板是不是在乎模型是否严格....
如果不在乎的话,
银行里面有一大堆聪明人(可惜不懂数学和统计)搞出来的quick & dirty的方法
保证让科班出身的人看了... 阅读全帖 |
|
w*********y 发帖数: 7895 | 6 我对3PL model的guessing value感的很困惑。是这样的,我用3PL里面得出的guessing
values 来检验不同的mixed 2PL and 3PL model. 比如说,for model 1,items with
guessing values greater than 0.3 was treated as 3PL model. For model 2,
items with guessing values greater than 0.2 was treated as 3PL model. For
model 1, items with guessing values greater than 0.1 was treated as 3PL
model. 我不知道这样做的话,最大的问题是什么。。。
我也做了点研究,找到一篇2012年的PAPER(HAN, 2012),是关于fixing c-
parameters. 他/她的观点是说,3PL 里面的guessing values 不reliable,而且和其他
2个parameters (a and b)是相... 阅读全帖 |
|
|
|
l****z 发帖数: 29846 | 9 October 31, 2012 by Warner Todd Huston
Based on his economic election model, Yale University economist Ray Fair
predicts that GOP nominee Mitt Romney will pull out a sparse victory in
November.
Fair’s model successfully predicts the winners of every election but two
since 1916. The two, the 1960 election of John Kennedy and the 1992 election
of Bill Clinton, are the only ones his economics-based electoral prediction
model gets wrong.
Fair uses several economic indicators in his model to determ... 阅读全帖 |
|
f******t 发帖数: 7283 | 10 因为话题重复,这里顺便也回复FangQ上面的帖子吧。既然你能把安全评分跟抗扭刚度
放在一起说,我比较确信你目前是不知道应该怎么去对一辆车子的抗扭刚度进行主观的
体验。
车身抗扭刚度跟被动安全性是两码事,分别测量车身结构两个不同的范畴。业界里这种
现象并不罕见:被动安全性很好,但抗扭刚度不行;当然也存在这样的车型,车身抗扭
刚度很好,但被动安全性不是太好。当然现在一线大厂的旗舰豪华轿车,都能做到被动
安全性和车身抗扭刚度都很好。Model S的抗扭刚度就是那个数值范围,这个数字肯定
不是无缘无故存在的。
具体来讲,Model S和Model X除了底部那个电池块,车身上部结构不是主要load
bearing部件(你不信的话可以跑去fremont工厂亲自问他们的设计师)。因此它们车身
的抗扭刚度主要由车底那块平板状的电池箱提供。你想想看,这么平薄的一块板子,跟
敞篷车类似,是没法提供足够的抵抗对角线扭力的能力的。
特斯拉的宣传向来都是对于有确凿优势的方面,必然大加宣传并且列出具体技术数据(
马力、加速测试等等);但对于车身抗扭刚度,你看他们怎么说?先不说给不给具体数
据了,甚至连“比xx竞... 阅读全帖 |
|
f******t 发帖数: 7283 | 11 类似的功率输出前提下,Model S的售价是跟full size旗舰豪华轿车(奔驰S550、宝马
750i、奥迪A8 4.0T)这些相当的。至于你说的Quattroporte,网上搜到的那个数据指
的是第5代,2003年推出的车型;而S80则是2006年推出的。咱们现在已经快进入2017年
了,把Model S跟十几年前的老车型比抗扭刚度,真是很不好意思的。
我们拿Model S跟现在的S550、750i、A8来比。你稍微做一下研究就能发现,这些车子
车身的抗扭刚度都是在40,000 nm/deg这个级别。换句话说,当前业界里面,这个级别
的轿车车身的抗扭刚度都在40,000 nm/deg这条线附近。这也是为何最新推出的Genesis
G90也要凑这个热闹,花大价钱把车身整到这个水平。长轴距轿车能把车身抗扭刚度做
到这个水平是很不容易的,大车厂情愿这样做那是因为这个指标对于实际驾乘感意义重
大。假如你已经长时间用过S/7/A8/LS这些车子,再跟Model S对比,除了加速的特性不
同之外,冷静思考感受之后你会发现Model S是全面大幅度落后的。
当然电车在加速上具有连贯性这个先天优... 阅读全帖 |
|
c********8 发帖数: 136 | 12 QuattroporteV 一直到2012年(第六代数据找不到, 但想必也不会高到哪去), S80现
在还是, model s本身也是2012推出的, 怎么能说是和十几年前的车型在对比? 还有
很多豪车的抗扭刚度根本没有公布, 所以并不能说model s这方面一定差。
我们还是拿genesis来做对比, 在价格低了几个级别的情况下, 它的抗扭刚度却比S/7
/A8还要高一些, 那是不是说这些车都很烂, 而神车XC90只有27000, 那是不是更弱
爆了?
所以说你还是不要再纠结抗扭刚度了, 事实就是这是个不重要的指标, 这也是为什么
基本上没有制造商会宣传这个数字, 即使他家这个数值再高, 因为大家都知道没几个
人care, 而你不停选择回避的安全评分才是大家都大力宣传的数据。 你去问问这个级
别车的买家, 有几个会说选择车的原因是抗扭刚度? 我相信扭刚度对你个人来说很重
要, 不然你也不会总纠结在这个问题上, 但是这总是少数, 对于一般人来说, 开平
路,高速(99%的时候)是根本感觉不到抗扭刚度的存在的
看到你说Model S是全面大幅度落后我就笑了, 除了你总在纠结的抗扭刚度, ... 阅读全帖 |
|
y******y 发帖数: 207 | 13 以下是关于Tesla model S 的安全性。虽然内容有些老旧,但大部分内容仍可以有效。
https://www.tesla.com/blog/tesla-model-s-achieves-best-safety-rating-any-car
-ever-tested
Independent testing by the National Highway Traffic Safety Administration (
NHTSA) has awarded the Tesla Model S a 5-star safety rating, not just
overall, but in every subcategory without exception. Approximately one
percent of all cars tested by the federal government achieve 5 stars across
the board. NHTSA does not publish a star rating above 5, however safety... 阅读全帖 |
|
|
c*****k 发帖数: 29 | 15 招聘Risk Modeler。工作地点曼哈顿。条件:理科,工科或经济学PhD,统计专业优先
。英文口语流利(
必备重要条件)。一年以上金融,建模或数据分析工作经验。
公司为Pre-IPO fintech company。预计两三年内上市,支持H1B。本组为新成立部门,
发展空间巨大。符合上述条件的申请人,请直接邮寄简历至c***[email protected]
As part of the Portfolio Management team, you will:
Own the testing and validation of statistical models used for important
business decisions and credit policies
Provide constructive feedback and recommendations for improvement to model
developers
Perform advanced qualitative and quantitative assessments of all aspects ... 阅读全帖 |
|
c*****k 发帖数: 29 | 16 招聘Risk Modeler。工作地点曼哈顿。条件:理科,工科或经济学PhD,统计专业优先
。英文口语流利(
必备重要条件)。一年以上金融,建模或数据分析工作经验。
公司为Pre-IPO fintech company。预计两三年内上市,支持H1B。本组为新成立部门,
发展空间巨大。符合上述条件的申请人,请直接邮寄简历至c***[email protected]
As part of the Portfolio Management team, you will:
Own the testing and validation of statistical models used for important
business decisions and credit policies
Provide constructive feedback and recommendations for improvement to model
developers
Perform advanced qualitative and quantitative assessments of all aspects ... 阅读全帖 |
|
|
m****s 发帖数: 18160 | 18 【 以下文字转载自 Statistics 讨论区 】
发信人: channelwin (Easonoo0), 信区: Statistics
标 题: Job opening in northern NJ- Statistical Modeler / Statistician
发信站: BBS 未名空间站 (Tue Feb 4 12:45:14 2014, 美东)
Statistical Modeler / Statistician needed for our firm. Lead development of
predictive modeling and analytics projects using data mining techniques (
including linear and logistic regression, neural networks, decision trees,
etc). Project and measure the effectiveness of modeling efforts and
marketing actions such... 阅读全帖 |
|
h****n 发帖数: 3447 | 19 Model 3的需求达不到让TSLA赚钱的量
就像Model S,虽然已经号称同级别的车里面销售的top 1,但是TSLA每年还需要亏很多
钱。说到底还是需求量不够。
Model S在豪华车里面,销售达不到让TSLA赚钱。Model 3在大众车里面,销售还是达不
到让TSLA赚钱。唯一的区别是卖的车在数目上增加了,还可以继续保持不赚钱是因为受
产量限制这个理由相当长的一段时间而已。但实际上既然能生产20万辆model 3,为啥
不能生产20万辆model S呢?相当于卖了60万辆的model 3。因为没有需求。 |
|
n****l 发帖数: 6652 | 20 不知你怎么算出$60000的均价的。
$35k 的base model, 加到$45k就是很不错的配置了。均价60k太离谱了。
:2018年第一季度,特斯拉共卖了21,815辆model S和model X。同时卖出了8,182辆
model 3。如果model S 和 model X的均价是95,000美元的话,那么model 3的均价大概
66,
:000美元。这已经是奔驰GLE350的价位了--如果不选4matic,差不多是出门价加上利息 |
|
发帖数: 1 | 21 不知道你这价格哪来的。model s便宜点的也就8万作用,其实现在的model 3 起步5万5
左右,价差刚刚好,可以吸引本来不愿意买更贵车型的。
我自己s和3都有,不存在鄙视链。我不买x的原因也很简单,它那个门的毛病听到太多
了,加上屁股不好看,我平常一人开也不喜欢太大个的车。
:所以Model 3不该存在,抢了自己生意还倒贴钱。最麻烦的是,本来Model S十几万起
,现在一群穷人买了几万的Model 3觉得这么便宜就高大上了,殊不知买Model S的人不
屑与买Model 3的为伍,好好的牌子给砸了。
: |
|
f*********e 发帖数: 8453 | 22 模特X看着就是一大号sedan。个人觉得设计上是失败的。为了酷炫弄那个门凭空增加成
本。
: 不知道你这价格哪来的。model s便宜点的也就8万作用,其实现在的model 3 起
步5万5
: 左右,价差刚刚好,可以吸引本来不愿意买更贵车型的。
: 我自己s和3都有,不存在鄙视链。我不买x的原因也很简单,它那个门的毛病听
到太多
: 了,加上屁股不好看,我平常一人开也不喜欢太大个的车。
: :所以Model 3不该存在,抢了自己生意还倒贴钱。最麻烦的是,本来Model S十
几万起
: ,现在一群穷人买了几万的Model 3觉得这么便宜就高大上了,殊不知买Model S
的人不
: 屑与买Model 3的为伍,好好的牌子给砸了。
: :
|
|
C*******s 发帖数: 330 | 23 出来冒个泡说说自己的经历。对于说Model 3不行的版友,尤其是下血本short的,建议
至少租一辆开两天再作判断。尤其建议hatwin试试,而不是天天一拍脑袋就是一个结论。
虽然之前在reservation list上,基于家里两辆车都是一年多新,而且看了一堆负面
review,几个月前排到的时候并没有打算买。考虑到有federal rebate,就在Turo租了
一辆开两天看看。一开以后就放不下手了。人说未来已经到了,只是来得不均匀,开完
Model 3就是这个感受。之前也开过Porsche什么的,虽然觉得不错,但绝对不会天天想
着,最多就是计划下辆车换。
Model 3的整车驾驶感受完爆所有开过的ICE,包括超跑。拿Model 3和Challenger作比
较就有点搞笑了。绝大部分人都不需要过分的加速,luxury的定义正在飞快地向智能化
驾乘感受转变,而传统车厂和Tesla的距离远不止三五年。现在的情况像极了iPhone刚
出来的时候。Model 3的autopilot 远超其它对手,使得长途驾驶和堵车变得异常轻松
,app的整合和方便让车变成像iPad一样的一个工具。我的一年... 阅读全帖 |
|
d***a 发帖数: 13752 | 24 他说的不对。本来讨论的是Model 3,他却去扯一个Model S过热的例子。
Model S是五六年前设计的老产品,那时电池技术并不成熟,发热也比较
厉害。今天的Model S,也还在用老电池组。就算是BBA,拿老电池组
来用,也会很头大。
今天这个时候,电池组发热和冷却,对电车来说已经不是大问题,对
Tesla来说也已经不是什么大问题。这是为什么我说他,半懂不懂还瞎扯。
https://www.teslarati.com/tesla-model-3-laguna-seca-raceway-track-racing-
video/
“The first session of the day had NO battery power limitations! Nine laps
without slowing down! Yes!”
The fact that the Model 3 was able to handle nine laps on a racetrack
without the car’s software limiting its performance is not... 阅读全帖 |
|
g******t 发帖数: 18158 | 25 2019-03-01 06:41 Seeking Alpha
+1.63%319.880特斯拉 (TSLA.US)
Tesla (NASDAQ:TSLA) has posted details on the launch of a $35K Model 3.
The standard Model 3 will feature 220 miles of driving range and a top speed
of 130 mph. The acceleration from 0 to 60 mph comes in at 5.6 seconds.
A second $37K version - called the Model 3 Standard Range Plus - will offer
240 miles of range, 140 mph as a top speed and 0-60 acceleration in 5.3
seconds.
Tesla confirms that sales are moving to online only all across ... 阅读全帖 |
|
s****e 发帖数: 1180 | 26 【 以下文字转载自 JobHunting 讨论区 】
发信人: sheide (shei), 信区: JobHunting
标 题: statistical score modeling, and scoring models 是什么东西? (转载)
发信站: BBS 未名空间站 (Fri Jul 1 18:28:30 2011, 美东)
发信人: sheide (shei), 信区: Statistics
标 题: statistical score modeling, and scoring models 是什么东西?
发信站: BBS 未名空间站 (Fri Jul 1 18:28:18 2011, 美东)
statistical score modeling, and scoring models 是什么东西?有一个职位,要面
这个东西,我google了一下没结果,哪位大大给解释一下,很感谢。 |
|
p***8 发帖数: 36 | 27 PhD and post-doc positions in Mechanical Engineering are immediately
available in the research group of Dr. Yi Wang at the University of South
Carolina-USC (Columbia/Main campus, https://www.sc.edu/study/colleges_
schools/engineering_and_computing/faculty-staff/yi_wang.php). USC is the
flagship university in the State of South Carolina, and the Ph.D. program at
the department of Mechanical Engineering is ranked No. 31 nationally by the
National Research Council (NRC) [1], and the College of Engi... 阅读全帖 |
|
s*******7 发帖数: 399 | 28 这个问题很好,因为这是离子通道voltage-sensing争来争去迄今尚无定论的地方。
先说题外话啊,哈哈。说到电生理,其实就是研究生物体如何把电信号转化为生理反应
。而电压依赖性离子通道(VGIC)是这个领域的研究热点之一,这个热点的核心又是
VGIC通道如何通过电压敏感区域(VSD)感受电压,VSD感受电压后又如何调控通道开放
?一句话,当外面来了电压,VSD在细胞膜间怎么运动?
在钾通道VSD结构没解析出来,有几种模型被提出来。
代表性的有2种:一种就是细胞膜相对静止,VSD的S4直上直下,即helical screw。另
一种VSD的S4相对静止,细胞膜直上直下,叫moving orifice。
这时期代表人物很多,我只记得Isacoff和Horn。Isacoff要提一下,一方面他是Lily
Jan的学生,另一方面他同时也是最早研制出TIRF的前沿人物之一。犹太人,很聪明。
如今研究VSD这一块实验室虽然很多,但基本上就出自几个大家族,诸如Miller C,
Lily,Catterall, Armstrong等。Mackinnon虽然名气大,但他和Yellen都出自Mille... 阅读全帖 |
|
b****a 发帖数: 352 | 29 我想试试蛋白 structural modeling。做 homology based modeling
39% sequence similarity
请问有这方面的经验的师兄师姐师弟师妹:
1, 39% sequence similarity 做结构 modeling ,可靠度如何?
2 ,就是想做个model 看看,预测几个residue 做做突变试试看。 最适合我入门,
learning curve 小的方法是什么。 我没有物理/编程/数学背景,做不来计算最低能
量什么的
我一个朋友推荐 modeller, 然后用pymol 看。 modeller 好像要用编程语言,我不是
academic lab ,购买/安装也要麻烦。 我看网上也有其他类似modeller 的平台。 如
果有经验的,希望推荐
谢谢 |
|
p***8 发帖数: 36 | 30 PhD and post-doc positions in Mechanical Engineering are immediately
available in the research group of Dr. Yi Wang at the University of South
Carolina-USC (Columbia/Main campus, https://www.sc.edu/study/colleges_
schools/engineering_and_computing/faculty-staff/yi_wang.php). USC is the
flagship university in the State of South Carolina, and the Ph.D. program at
the department of Mechanical Engineering is ranked No. 31 nationally by the
National Research Council (NRC) [1], and the College of Engi... 阅读全帖 |
|
p***8 发帖数: 36 | 31 PhD and post-doc positions in Mechanical Engineering are immediately
available in the research group of Dr. Yi Wang at the University of South
Carolina-USC (Columbia/Main campus, https://www.sc.edu/study/colleges_
schools/engineering_and_computing/faculty-staff/yi_wang.php). USC is the
flagship university in the State of South Carolina, and the Ph.D. program at
the department of Mechanical Engineering is ranked No. 31 nationally by the
National Research Council (NRC) [1], and the College of Engi... 阅读全帖 |
|
p***8 发帖数: 36 | 32 PhD and post-doc positions in Mechanical Engineering are immediately
available in the research group of Dr. Yi Wang at the University of South
Carolina-USC (Columbia/Main campus, https://www.sc.edu/study/colleges_
schools/engineering_and_computing/faculty-staff/yi_wang.php). USC is the
flagship university in the State of South Carolina, and the Ph.D. program at
the department of Mechanical Engineering is ranked No. 31 nationally by the
National Research Council (NRC) [1], and the College of Engi... 阅读全帖 |
|
p***8 发帖数: 36 | 33 PhD and post-doc positions in Mechanical Engineering are immediately
available in the research group of Dr. Yi Wang at the University of South
Carolina-USC (Columbia/Main campus, https://www.sc.edu/study/colleges_
schools/engineering_and_computing/faculty-staff/yi_wang.php). USC is the
flagship university in the State of South Carolina, and the Ph.D. program at
the department of Mechanical Engineering is ranked No. 31 nationally by the
National Research Council (NRC) [1], and the College of Engi... 阅读全帖 |
|
p***8 发帖数: 36 | 34 PhD and post-doc positions in Mechanical Engineering are immediately
available in the research group of Dr. Yi Wang at the University of South
Carolina-USC (Columbia/Main campus, https://www.sc.edu/study/colleges_
schools/engineering_and_computing/faculty-staff/yi_wang.php). USC is the
flagship university in the State of South Carolina, and the Ph.D. program at
the department of Mechanical Engineering is ranked No. 31 nationally by the
National Research Council (NRC) [1], and the College of Engi... 阅读全帖 |
|
p***8 发帖数: 36 | 35 PhD and post-doc positions in Mechanical Engineering are immediately
available in the research group of Dr. Yi Wang at the University of South
Carolina-USC (Columbia/Main campus, https://www.sc.edu/study/colleges_
schools/engineering_and_computing/faculty-staff/yi_wang.php). USC is the
flagship university in the State of South Carolina, and the Ph.D. program at
the department of Mechanical Engineering is ranked No. 31 nationally by the
National Research Council (NRC) [1], and the College of Engi... 阅读全帖 |
|
p***8 发帖数: 36 | 36 PhD and post-doc positions in Mechanical Engineering are immediately
available in the research group of Dr. Yi Wang at the University of South
Carolina-USC (Columbia/Main campus, https://www.sc.edu/study/colleges_
schools/engineering_and_computing/faculty-staff/yi_wang.php). USC is the
flagship university in the State of South Carolina, and the Ph.D. program at
the department of Mechanical Engineering is ranked No. 31 nationally by the
National Research Council (NRC) [1], and the College of Engi... 阅读全帖 |
|
s**k 发帖数: 10 | 37 Modeling without Programming is a systematic approach developed by Lin Chen
to implement advanced algorithms and models in financial engineering,
computational finance and actuarial science in Excel spreadsheets using
Excel’s functions only, without VBA coding.
Such advanced models and methods include, but are not limited to, HJM model,
BGM model,Monte Carlo method, least-square MC, American option pricing,
extreme value theory, copula approach, default risk models. These models and
algorithms a |
|
Y******u 发帖数: 1912 | 38 Total comp $547,000 as a computer system analyst with 3 year working
experience? Two sigma is so generous
http://www.forbes.com/sites/nathanvardi/2014/02/19/analyst-char
Kang Gao, a 28-year-old who until recently worked as a quantitative analyst
at Two Sigma Investments, was arrested last week and charged by the
Manhattan District Attorney of felony offenses, including computer trespass,
criminal possession of computer related material and unauthorized use of
secret scientific material. Gao admi... 阅读全帖 |
|
Y******u 发帖数: 1912 | 39 Total comp $547,000 as a computer system analyst with 3 year working
experience? Two sigma is so generous
http://www.forbes.com/sites/nathanvardi/2014/02/19/analyst-char
Kang Gao, a 28-year-old who until recently worked as a quantitative analyst
at Two Sigma Investments, was arrested last week and charged by the
Manhattan District Attorney of felony offenses, including computer trespass,
criminal possession of computer related material and unauthorized use of
secret scientific material. Gao admi... 阅读全帖 |
|
|
M*********m 发帖数: 2024 | 41 【 以下文字转载自 Physics 讨论区 】
发信人: Microsystem (clam), 信区: Physics
标 题: 从二维 Ising model 说起
发信站: BBS 未名空间站 (Tue Dec 18 00:06:02 2007)
呵呵,我一个人是写不完的。我们从这个问题开始,来讨论一下固体的磁性,每人负责
一个小问题,尽量在这周里做一番研究,周末的时候讨论一下,怎么样?
下面有几个问题:
1. 物质的各种磁性,Ising model 和 Heisenberg model 的关系
2. 2D Ising model和自发磁化的矩阵解
3. 各向同性的经典 Heisenberg model 和 Wagner-Mermin 定理
4. 磁体的元激发及其量子化
5. t-J model,hubbard model 的简介
6. 二维超导体的反铁磁相(superstring) |
|
l*******y 发帖数: 22 | 42 几点想法
第一:不明白为什么属于bayesian的hierarchical linear model会和confidence
interval这样的名词联系在一起,应该是
posterior interval吧?还有,啥是post hoc power analysis?就gelman那本书:
bayesian data analysis里面讲到的
hierarchical linear model而言,主要的目标是计算parameter的posterior
distributions一类的统计量,而之所以用
hierarchical linear model,很多情况下是因为这种模型可以combine various data
sources,并且在一个linear model
中加入其他的random terms (compare to the error term in linear model), 有助
于model the extra variance
components. 所以hierarchical linear model的基本用途是这么回事。当然,
hierarchi |
|
y*****w 发帖数: 1350 | 43 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Not necessarily. Whether这个model找到的候选基因是和体重相关的 or 是和抗药相
关的 depends on what the output of the model shows:
1) If the main effect of基因表达(fixed effect) is significant, it means找到
的候选基因是和体重相关的.
2) If the interaction term of基因表达×用药(fixed)has a significant effect
, then 找到的候选基因是和抗药相关的.
3) If both the main factor of 基因表达 and the interaction term of 基因表达
×用药 have a significant effect, then 找到的候选基因是和 both 体重 and 抗药
相关的.
Overall, this is a r... 阅读全帖 |
|
a*****i 发帖数: 1045 | 44 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 |
|
w*******y 发帖数: 60932 | 45 I was just at Lowes, and they had a few left over Weber Q 120's for $89,
regular $169. These are the older models. Item number is 297992, model
number 516002. The newer version of the same grill is model 516501, and
retails for $179. The model number of either type was printed on stickers
that are visible from any of the 4 sides, so you could easily check top
stock to see if they have some of the older model left over. (If website
says limited qty).
Remember to use a Lowes card, $10 off 50 or 10... 阅读全帖 |
|
w*******y 发帖数: 60932 | 46 I was just at Lowes, and they had a few left over Weber Q 120's for $89,
regular $169. These are the older models. Item number is 297992, model
number 516002. The newer version of the same grill is model 516501, and
retails for $179. The model number of either type was printed on stickers
that are visible from any of the 4 sides, so you could easily check top
stock to see if they have some of the older model left over. (If website
says limited qty).
Remember to use a Lowes card, $10 off 50 or 10... 阅读全帖 |
|
c*******9 发帖数: 9032 | 47 新浪科技讯 北京时间10月5日早间消息,特斯拉(180.98, 7.67, 4.43%)(TSLA)CEO艾
伦-马斯克(Elon Musk)周五发表博客文章解释Model S起火事件,成功安抚投资者并促
使该公司股价反弹4.43%,收于180.98美元。
视频:特斯拉Model S电动汽车因发生撞击着火
媒体来源:新浪科技
这是教科书一般的危机公关案例。
事件缘起于周三,一辆特斯拉Model S起火的视频流传网络,点燃了公众及投资者
对于电动汽车安全性的担忧。特斯拉股价自周三下午视频传出后迅速下跌,并在第二天
继续下探。周三和周四两个交易日,特斯拉股价由192.99美元跌至173.23美元,跌幅为
10.24%,或19.76美元,市值损失约24亿美元。
尽管特斯拉Model S型电动车曾在安全碰撞测试中获得美国公路交通安全管理局(
NHTSA)的最高分5.4星,公众仍然对于电动车这一相对新生事物的安全性存有一定程度
的怀疑。投资者也普遍认为,潜在的重大安全事故将对腾飞了半年多的特斯拉股价造成
冲击。2013年3月中旬的时候,特斯拉股价还在35美元上下徘徊,2013年9月底,该股便
创... 阅读全帖 |
|
f******t 发帖数: 7283 | 48 Model S, Model X的结构强度只能说是一般。
Model S的small overlap结果就能看出来,脚部空间形变是28,22cm等这种量级,相当
差劲的。相比之下连Camry都比它好上个10倍(实测结果是2-4cm)。
把IIHS各项实测参数对比一下就知道,连Camry都能完胜Model S。至于结构更弱的
Model X那就更没法比啦。
尝鲜欲望是到底强烈到什么程度才会使人丧失理性思维买tesla呢?还好幸亏保时捷
Mission E马上开卖了,保时捷的品质同样的售价让你两者择一,谁还会买半成品般的
model s呢。
X
car |
|
|
C**T 发帖数: 5379 | 50 昨天去新开的一家Tesla体验店仔细的看了看Model S,特意关注了装配的细节,发现
Model S在装配上确实和同档次的德车日车相差甚远,倒是很符合美国车的特点,就是
大条,不细致,不精益求精。
Model S的设计还是很赞的,可是仔细看,用料和装配质量上有很大的问题。举几个简
单的例子:
1)前后车窗的银色frame接缝处对的不齐。店里两辆车,四个解封可谓远近高低各不同。
2)前trunk和左右fender的接缝宽窄不一,甚至有一边靠近前保险杠处还漏出了里面的
防水橡胶。
3)车门一打开,好家伙,一大块黑乎乎的板,看起来好像是皮的,摸起来手感却象塑
料。
4)皮dashboard,摸起来也象塑料。
5)stitches摸起来特别硬,有点划手。
6)中控那个大触摸屏下面什么都没有,空荡荡的,开车时如果把胳膊搭在中间的储物
盒上方,会感觉胳膊要掉下去。
7)座椅的真皮看起来还不如2万出头的Accord。
8)底盘焊接工艺看起来比较糙。
当然,Model S的优点众多,就不总结了。总体来说,对于8万-10万的车,这样粗糙的
装配质量是很难让人信服的。同样级别的S系,7系,A8,LS,做... 阅读全帖 |
|