C.J. DUAN

Data Scientist, Brainbench CFA, Ph.D

[ adbs3p@r.postjobfree.com Ó 757-***-**** R Blacksburg, VA 24060 ® www.dulun.com/

github.com/hublun

WORK PHILOSOPHY

“The vlaue of a model lies in the quality

of its predictions”

Thomas W. Miller

MOST PROUD OF

3

Servire Est Vivere

helping Braydon Farm LLC in Alabama

optimize their daily flower delivery

routes using Transportation Model

I

Original Research

from Business Process Outsourcing to

Bayesian Network, 5 journal articles in

5 years in a row

3

Machine Learning

from Random Forests to Convolu-

tional Neural Network

Certified Financial Analyst

5/6/2000 Brainbench transcript ID

565042

STRENGTHS

Bayesian Analysis with rStan, RStudio

Scala & Stanford Core NLP

Python & NLTK &Word2Vec & PyTorch

SAS Studio & Enterprise Miner, RapidMiner

EDUCATION

Ph.D in Industrial Management

Clemson University

2000 – 2007

MBA, Finance

Morgan State University

1998 – 2000

B.Eng. in Optical Instruments

Tsinghua University

1986 – 1991

EXPERIENCE

Assistant Professor of Quantitative Methods

Troy University

March 2009 – May 2017 Norfolk, VA and Troy, Al Engineer

EAC

Oct. 1992 – July 1997 Beijing, China

PROJECTS

Page Recommendation Model based on NLP + PyTorch with Alternative Bayesian Implementation

One systemwithtwoimplementations

Jan. 2018 - Dec. 2019 Computing Project

Using (1122)users’ browsing data of (731) web pages of one website), we derived first a 1122 X 731 matrix of user-page interaction frequen- cies. Second, we applied the Word2Vec algorithm in the Gensim pack- age to obtain embeddings of all keywords associated with each page. The embeddings were further combined using the Doc2Vec algorithm to create one 222-dimension vector for each webpage. The two matri- ces were then fed as input and real values into a one-layer Deep Learn- ing network constructed using PyTorch 1.3.1. After training, the model yielded accuracy of 24 in terms of MSE loss.

Alternatively, we developed a Bayesian Log-Normal model in Stan 2.19.1 and run the model using MCMC simulation with the same two matrices as inputs and ground truth. We assessed the same MSE and obtained a final value of around 35.

Bayesian Data Analysis (BDA) of Home Field Advantage together with Ananyo Chakravarty (Troy University MBA 2017)

Jan. 2016 - 2018 Research Project

We developed our own data scraper (in Python) and analyzed our col- lected data set with bnlearning package in R. The result is again another research paper (Bayesian Learning from Top European Professional Soc- cer Leagues about Attendance Influencers: An Examination at the Sea- son Level).

Phishing Websites Classification

with Anuj Gurav (Troy University MBA 2016)

Jan. 2015 - 2017 Research Project

We perform RandomForest analysis (in RapidMiner and R) of Phishing website data and produced a research paper (Exposing the Type I and II Errors Made by Machine Learning Algorithms: Revisiting the boy who cried wolf in the context of phishing detection).

REFEREES

Dr. Ronald Shehane

[ adbs3p@r.postjobfree.com

R Troy University

Dr. Ingyu Lee

[ adbs3p@r.postjobfree.com

R Troy University

Dr. William Neese

[ adbs3p@r.postjobfree.com

R Troy University

REMARKS

". . .On behalf of my wife and I, we would like to sin- cerely thank Dr. Duan, Mrs. Amy Hu, and you for the time you spent developing a distribution plan for our new floral operation. We have incorporated it into our operation and it appears to have significantly reduced our travel time and expenses in distribution our prod- ucts to the Marvin Stores chain. Hopefully this will be an on-going study and we will be able to “tweak” or al- ter the system based on your recommendations as our business grows. . ."

Stephen (Brad) Bradshaw Garrott

Owner, Braydon Farms, 2015

". . .I work on projects focused on educating high school students in Mississippi in entrepreneurship opportu- nities and what resources are available to start-ups. I use Excel frequently and I am thankful for being familiar with excel formula’s like vlookup and IF. ... I wish there was an Operations Management class because that

was my favorite in undergrad and where I find myself passionate. I am considering entering the workforce aiming for an Operations Management position based on my success in your class and how much I enjoyed it. I think your class is very important and I’m glad I had the opportunity to learn from you. I am also open to entrepreneurship opportunities as I am so active in the entrepreneurship . . ."

Russell, QM 3345, 2016

PUBLICATIONS

> Under Review

Duan, C.J. (2019). Home Sweet Home-Field Advantage in Professional Soccer: A Multilevel Bayesian Analysis. Revision submitted on 6/16/2019, Journal of Business Analytics.

– (2018a). Exposing Model Bias in Machine Learning: Revisiting the Boy Who Cried Wolf in the Context of Phishing Detection. submitted to Infor- mation Systems and Operations Research.

– (2018b). Retrieving Hypotheses Parameters from Real Life Events: The Case of Teaching Type I and II Errors in Business Statistics Course. submit- ted to Asian Journal of Management Cases.

q Journal Articles

Duan, C.J. and Graham S. Pierce (2017). “Starting Operations Man- agement with a ‘Big Bang’: Using Sitcom to Introduce OM Concepts to Students”. In: Operations Management Education Review 11, pp. 115– 122.

Duan, CJ, J Hu, SC Garrott, et al. (2016). “Using Excel Solver to Solve Braydon Farms’ Truck Routing Problem: A Case Study”. In: South Asian Journal of Management Sciences (SAJMS) 10.1, pp. 38–47. DOI: 10.21621/sajms.2016101.04.

Teng, Joe KL et al. (2016). “Determinants of Research and Develop- ment Intensity from a Network Perspective”. In: Studies in Business and Economics 11.3, pp. 129–139. DOI: 10.1515/sbe-2016-0040.

Duan, Chaojie, Varun Grover, Nicholas Roberts, et al. (2014). “Firm valuation effects of the decision to adopt relationally governed busi- ness process outsourcing arrangements”. In: International Journal of Production Research 52.15, pp. 4673–4694. DOI: 10.1080/00207543. 2014.884289. eprint: https://doi.org/10.1080/00207543.2014. 884289. URL: https://doi.org/10.1080/00207543.2014.884289.

Kronenburg, Maria A et al. (2013c). “The Impact of Contract Design on Contract Performance Satisfaction”. In: International Journal of Management Research and Reviews 3.4, p. 2673.

Duan, Chaojie, Varun Grover, and Nagraj (Raju) Balakrishnan (2009).

“Business Process Outsourcing: an event study on the nature of pro- cesses and firm valuation”. In: European Journal of Information Systems 18.5, pp. 442–457. ISSN: 1476-9344. DOI: 10.1057/ejis.2009.38.

Conference Proceedings

Duan, C.J. (2013). “Retrieving Hypothesis Parameters From Real Life Events”. In: The 43rd

Southeast Decison Sciences Institute Conference

Proceedings. SE DSI.

Kronenburg, Maria A et al. (2013a). “Public Management Skills Needed in Contract Development”. In: The Southeastern Conference for Public Administration. SECoPA.

– (2013b). “Public Management Skills Needed in Contract Develop- ment”. In: The Southeastern Conference for Public Administration. SEC- oPA.

– (2012). “The Impact of Contract Design on Contractor Performance

- A Second Look”. In: 5th

International Public Procurement Conference.

IPPC.

Duan, C.J. (2004). “Teams in TQM: an Knowledge based View”. In: The 35th

DSI.

– (2001). “One Machine Scheduling of Jobs to Minimize Total Weighted Tardiness”. In: The 37th

Southeast INFORMS. SE INFORMS.

Contact this candidate