c******s
发帖数: 18 | 1 地点：Cincinnati, OH
首先说明，我没有招Ph.D.的budget。您要是Ph.D.，麻烦您先私信通知一下您的salary
expectation。请报一个数，请不要报一个range,也请不要说open。
sponsor绿卡，H-1。但是如果您现在是cap-exempt的H-1，那对不起了。不支持eVerify
。您要打算用29个月OPT，那只好不谈。
您有兴趣的话请把简历用私信发这里。这样不成的话大家也好说话。格式乱就乱
要求，最重要的是statistics，要求generalized linear mixed effect model;
您要说您懂的话那请详细说说都用过什么variance structures,为什么用。要求懂MCMC;
您要说您懂的话请详细说说都写过什么likelihood,为什么。
其次请详细说一下您理解的整个modeling的过程，从有data开始到deploy the model
for production
对machine learning algorithm也有要求。您懂的话请说说您作过什么，为什么用那些
models
对SQL，simulation, optimization,VBA等等也有要求不过比较次要 | m*********n
发帖数: 119 | | c******s
发帖数: 18 | 3 不支持eVerify
【在 m*********n 的大作中提到】
: 您要打算用29个月OPT，那只好不谈。
: Y?
| s********1
发帖数: 54 | 4 ______________________________________________________________________
In terms of the variance structure
______________________________________________________________________
Normal structure depend on each individual,Spatial structure depends on the
distances between two points and compound, etc.
______________________________________________________________________
In terms of the likelihood:
______________________________________________________________________
The PL method is based on Wolfinger and O'Connell (1993) who proposed an
algorithm to calculate the parameter estimates and the values of fixed and
random effects. The main idea of this algorithm is to approximate the GLMM
as a LMM through use of a pseudo-variable obtained through a Taylor series
expansion. The algorithm iterates between updates of the pseudo-variable and
parameter estimator that result from the LMM computations.
The Laplace method is based on Booth and Hobert (1998) who utilized a
Laplace approximation based on the work of de Bruijn (1981) to approximate
the value of the BP. An iterative strategy is employed to obtain the eBP,
first approximating it using current values of parameters and fixed effects,
and then updating those values by approximating the likelihood using
another Laplace approximation. This process continues until convergence is
achieved. The precision of the eBP is evaluated using a Taylor series
approximation to the conditional mean squared error (CMSE) derived in Booth
and Hobert (1998). Zhao et al. (2006) and Skrondal and Rabe-Hesketh (2009)
have also advocated CMSE as a suitable measure of precision.
The quadrature method also calculates the BP using a Laplace approximation,
however the likelihood function is approximated by an adaptive quadrature
approximation [see, for example, Golub and Welsch (1969), Abramowitz and
Stegun (1972) and Pinheiro and Chao (2006)]. The advantage of the adaptive
quadrature approximation is to improve the approximation of the likelihood
function by centering and scaling the quadrature points. Again, the same
iterative strategy as the Laplace method is employed until convergence
criteria is met and CMSE is used to measure the precision of the eBP. It is
worth noting that the CMSE is a function of parameter estimates, and thus
the value of CMSE is not the same for the Laplace and quadrature methods
since the plug-in parameter estimates are calculated by a Laplace and an
adaptive quadrature approximation, respectively.
______________________________
In terms of 从有data开始到deploy the model
________________________________________
It depends on the data. If this data is countable, I usually consider the
poisson GLMM or negative binomial. And when I deploy the model, I would
start to compare the likelihood and also observe the significant covariates
in the GLMM.
______________________________________________________
To summary, I have studied GLMM for 4 years. I am very familiar to this
model. If you want to get any further information regarding to this model,
please feel free to contact me at 7076240634.
Cheng-Hsueh Yang
My resume is as follows:
Cheng-Hsueh Yang
Department of Statistics Phone: (707)
6240634
University of California Email: cyang007@
ucr.edu
1108 Olmsted Building
Riverside, CA 92507
Education
University of California: Riverside, CA
Ph.D. in Statistics, December 2012
Graduate Fellowship
National Central University (NCU): Chung-Li, Taiwan
M.S. in Statistics – September 2006
Certifications and Software skills
•SAS Certified Base Programmer for SAS 9 credential
•SAS Certified Advanced Programmer for SAS 9 credential
•SAS, SQL, Stata, Minitab, R, SPSS, Splus, Matlab, Latex and Excel.
Project Experience
• Revenue Forecasting
Methodology: Regression, nonlinear, linear mixed model and principle
component analysis
Determined how region, number of employees, salary, and interest rate
influence business income potential
• Graduate School Admission
Methodology: Logistic model, probit model and K-means clustering
Determined how age, gender, race, GPA and GRE influence the admission
potential
• Banking Cost Minimization
Methodology: Ad-Hoc analysis and quantitative analyses
Proposed an optimized strategy to decrease cost using commercial and
customer information
• Study of Diabetes Patients
Methodology: Kaplan-Meier analysis, Cox PH models, AFT models and
parametric survival models
Determined how age, gender, race, treatments and economics influence the
death of diabetes potential
• Study Ranking Pharmaceutical Company
Methodology: Nonparametric methods including sign test, kruskal wallis
test and nonparametric regression
Determined how treatments and number of the researchers influence the
ranking potential
• Study of Recovery from Surgery
Methodology: Longitudinal and generalized linear mixed model
Determined how treatments, gender, area, and the equipment ranking
influence recovery days potential
• Prediction of Stock Prices
Methodology: the Metropolis-Hastings, Monte Carlo, MCMC or EM algorithm
Proposed a more reliable algorithm to predict the market value
• Complete Experimental Design
Methodology: Survey design, experimental design including block, RCBD,
factorial and fractional designs
Provided students’ opinions about school relocation with a quantitative
and qualitative report
Research Experience
Prediction Intervals for generalized linear mixed models 2012
University of California, Riverside,
•Research Advisor: Dr. Daniel Jeske
•Derived the general prediction interval by using the best linear
predictor in generalized linear mixed
models
Tests for Right Censored Paired Survival Data 2005 - 2006
National Central University
•Research Advisor: Dr. Yuh-Ing Chen
•Improved the logrank test and the test predicting the difference of
two restricted mean
survival times in paired right censored data
Work Experience
National Taiwan University College of Medicine 2006
•Assisted a medical researcher in collecting data and analyzed the
distribution of a virus
Department of Statistics in NTPU 2005 - 2006
•Assisted an instructor in preparing lectures and graded homework
HandsOnNetwork 2004
•Assisted the company for designing a survey about relocating a
school
A.C.Neilson Ltd. 2004
•Assisted the company with analyzing surveys such as estimating the
number of households who own a TV and evaluating the water quality in Taiwan
Presentations
American Statistical Association: Joint Statistical Meetings
•Different predictors for the Generalized Linear Mixed Model 2010
Studied the performance of the convergence of different predictors
•Prediction Intervals with the BLP for Generalized Linear Mixed
Models 2011
Constructed a prediction interval based on the best linear predictor and
evaluated the coverage probability and average length
•General Prediction Intervals for Generalized Linear Mixed Models 2012
Generalized the prediction interval and improved the problem of
inconsistency for the best linear predictor by using an appropriate
transformation
American Statistical Association : Quality and Productivity Research
Conference
•Prediction Intervals for Generalized Linear Mixed Models 2012
Demonstrated some examples with the transformation of the best linear
predictor and compared them with previous methods such as the pseudo-
likelihood based and conditional mean squared error based methods
Teaching experience
Teaching Assistantships: UCR 2008 - 2012
•STAT100A: Introduction to Statistics
Lead labs using Minitab, graded homework, and lead discussion sessions
covering topics such probability, linear regression models, distributions,
hypothesis testing, confidence intervals, etc
•STAT48: Statistics for Business
Lead labs using Excel, graded homework, and lead discussion sessions
covering topics such as probability, linear regression models, distributions
, hypothesis, confidence intervals, etc
•STAT100B: Introduction to Statistics
Graded students’ quizzes, and lead discussion sessions covering topics such
as linear regression models, analysis of variance, nonparametric methods,
simple experimental designs, etc
Lecturer in UCR 2011
•STAT100A: Introduction to Statistics
Instructed students covering topics such as descriptive statistics, linear
regression models, distributions, hypothesis testing, confidence intervals
and created labs and quiz materials
salary
方案
MCMC;
【在 c******s 的大作中提到】
: 地点：Cincinnati, OH
: 首先说明，我没有招Ph.D.的budget。您要是Ph.D.，麻烦您先私信通知一下您的salary
: expectation。请报一个数，请不要报一个range,也请不要说open。
: sponsor绿卡，H-1。但是如果您现在是cap-exempt的H-1，那对不起了。不支持eVerify
: 。您要打算用29个月OPT，那只好不谈。
: 您有兴趣的话请把简历用私信发这里。这样不成的话大家也好说话。格式乱就乱
: 要求，最重要的是statistics，要求generalized linear mixed effect model;
: 您要说您懂的话那请详细说说都用过什么variance structures,为什么用。要求懂MCMC;
: 您要说您懂的话请详细说说都写过什么likelihood,为什么。
: 其次请详细说一下您理解的整个modeling的过程，从有data开始到deploy the model
| s********1
发帖数: 54 | 5 When I sent you the message, I got the following error message.
critmass：发送失败(发信失败)！
Can you provide your email to me? Thanks!!! | s*********e
发帖数: 25 | 6 因为人家不是E-verify的公司呀，OPT只让用12个月
【在 m*********n 的大作中提到】
: 您要打算用29个月OPT，那只好不谈。
: Y?
|
|
