Experienced statistical modeling and analytics

Liping Deng

Phone Number: 937-***-****

Email: *********@*****.***

Address: **** ******** ***** **.

Marietta, GA, 30062

Profile

Experienced statistical modeling and analytics professional with expertise in predictive models, explanatory data analysis and data mining solutions

Proficiency in modeling and analysis tools: SAS E-Guide, SAS E-Miner, R, SQL, EXCEL

Strong finance background in risk management modeling and marketing analytics

Related Experience

Analyst - Modeling

GE Capital Oct 2014 – Present

Developed and monitored origination PD model for use in risk management in loan and leasing finance services

-Developed reject inference fusion model to define “bad/good” for non-booked potential customers

-Monitored external and in-house data in quality and distribution stability

-Leveraged decision tree variables in modeling

-Conducted segmentation analysis to improve model

-Developed multi-bureau PD model and built a standard modeling procedure

-Completed model sensitivity analysis

Completed challenger models of multi-bureau origination PD model

-Compared different methods including traditional decision tree, bagging, gradient boosting and neural networks, and built decision tree, which is comparable with champion model in terms of model performance, robustness and implementation

Sr. Statistical Analyst

CAN Capital, Inc. Feb 2014 – Oct 2014

Achieved 31% campaign lift by building logistic regression model to predict the probability of customer application

Defined high response prospects and key customer attributes by conducting decision tree analysis and market segmentation analysis

Identified peak industries that vary with month in a year by conducting seasonality analysis at the SIC2 (Standard Industrial Classification) level

Identified key features and increased the effectiveness of direct mail campaign by performing multivariate experimental design

Statistician

Credigy LLC Apr 2013 – Jan 2014

Predicted debtor default risk and optimized returns by building logistic regression model

Completed model validation and created performance reports

Helped make bid decisions on debt packages by predicting charge-off amounts per month

Operations Research Analyst

Rainmaker Group LLC Apr 2011 – Jul 2012

Optimized the dynamic rate of hotel rooms for group customers by applying segmentation analysis and logistic regression models

Conducted seasonality analysis over hotel price using the Fourier series

Forecasted hotel gambling revenue by employing double exponential model and linear regression model

Statistical Analyst (Contractor)

Providence, LLC. Apr 2009 – Apr 2011

Characterized the potential sources of air pollution using factor analysis and time series analysis

Research Assistant

Applied Statistics Program, Wright State University May 2008 – Mar 2009

Investigated placebo effect versus anti-depression drug effect using regression model, meta-analysis, and bootstrapping technique

Education

Master in Applied Statistics GPA: 4.0/4.0

Wright State University, Department of Mathematics and Statistics

Master of Business Administration Concentration in Finance GPA: 3.8/4.0

Wright State University, Raj Soin College of Business

Bachelor of Engineering in Landscape Architecture GPA: 3.0/4.0

Chongqing University, China, Architecture Department

Certificates

SAS/Macro

Advance Analytics for Customer Intelligence Using SAS

Publication

L. Deng, et al, “Interpreting Meta-Regression: Application to Recent Controversies in Antidepressants' Efficacy”, Statistics in Medicine, featured article, Vol 32, No 17, Jul 2013.

Contact this candidate