Fei 'Kevin' LU, Ph.D

Dallas, TX ***** 972-***-**** ac745z@r.postjobfree.com

CAREER SUMMARY

Analytics professional with 10 years experience in providing predictive modeling, risk/fraud analysis, marketing research and quantitatively-based consulting to address strategic business issues. Demonstrated success in defining and delivering actionable, business objective-focused analytic solutions through:

• Unique combination of analytic skills, marketing / business acumen and technological knowledge.

• Proficiency in applying advanced data mining and machine learning methodology & Algorithms in various industries such as banking, real estate, mortgage, telecommunication, retail, insurance, loan product and etc.

• Strong practical skills of modeling/analytic tools and their application to business as well as principles of financial & credit risk management analysis.

• Successfully developed advanced statistical & economical models, algorithms, and tools in a wide range of technical and business area including many areas of Research & Modeling, Risk/Fraud Analysis, Advertising, Marketing Strategy,B2C, B2B, and more.

• Passionate professional focus is developing cross-vertical practices including marketing-mix optimization, segmentation, customer base analytics, analytic consulting and market research. CORE ANALYTICS SKILLS

PROFESSIONAL EXPERIENCE

Principal, Property Intelligence & Risk Management Solution, (Jan 2017 – Nov 2018) Corelogic Inc, Irving, TX USA

Perform comprehensive housing market & residential mortgage data analysis.

Responsible to maintain and oversee the mortgage fraud index model based on residential mortgage loan applications processed by CoreLogic LoanSafe Fraud ManagerTM, which provide key metrics and inputs to the CoreLogic Mortgage Fraud Report that analyzes the collective level of loan application fraud risk experienced in the mortgage industry each quarter. The mortgage fraud report is to monitor the fraud risk trend in the real estate market.

Lead the development, coding, implementation and documentation of a new nationwide Housing Market Data-driven Multi-layer Segmentation model to create parcel-level segmentation & classification strategy using consolidated household-centric Bureau data such as Experian Gold to Corelogic property-centric Datamart.

Filed a patent on the new approach called “Distinction Index” to search the optimal segment number so that the profiling performance of the housing segmentation strategy can be optimized and enhanced.

Built quarterly fraud trend forecasting of the residential mortgage applications based on the fraud score range cross market segments (ZIP code, county, CBSA, top 100 U.S. metro areas, nation) using ARIMAX technique and calibrate the secondary market (CapMarket) mortgage fraud scores.

Monitored the performance of P2List and P2Purchase models (Purchase, HELOC and Refinance), performed model validation and assessment and provided advice to business management and identified opportunities to apply quantitative methods to improve model performance. Used SAS/macro and excel to automate the validation process and build comprehensive assessment report.

Predictive/Prescriptive Modeling

• OLS/logistic regression

• Survival analysis

• Data mining methods

• Time series(Box-Jenkins) model

• PCA, FA and Cluster Analysis

• Experiment design

• Machine Learning Algorithms

Credit, Risk & Fraud Analysis

• Fraud detection analysis

• Credit Risk Scorecard Modeling

• CCAR modeling

• Benchmarking

• Backtesting

• CECL Forecasting(PD,LGD,EAD)

• Stress Testing

Marketing Analytics

• Direct & Database marketing

• CRM analysis

• Customer acquisition

• Customer retention & loyalty

• Marketing-Mix Model (MMM)

• Campaign analysis

• Marketing Strategy Optimization

Sr. Statistical Modeler, Department of Analytics (Sept 2015 – Sept, 2016) LexisNexis Risk Solution,Atlanta, GA USA

Performed cross-industry insurance fraud detection and built fraudulent scorecard on auto insurance client’s data using advanced statistical techniques.

Conducted semi-year validation on LexisNexis Risk Classifier model that utilizes data from attributes derived from public records, driving history and credit to help better assess a proposed insured’s risk profile – then distilled it into a numeric score with reason codes in auto insurance underwriting and rating processes.

Maintained nationwide auto insurance fraudulent database, perform customer testing and ad hoc analysis.

Built auto-insurance industry fraudulent trend forecasting model & quarterly report using time series techniques. Sr. Statistician, Data Science Division (May 2013 – Aug 2014) Valuescope Inc., Southlake, Texas, USA

Performed quantitative analysis including customer profiling, marketing-mix optimization, customer churn model to enhance the marketing efficiency for auto-dealership and newspaper clients.

Managed, participate in a wide variety of planning and research and complete projects with the defined schedules.

Independently identified new business opportunities, and collaborated with division president to solve existing business problems for clients using cutting-edge quantitative methods. Manager/Sr. Data Analyst, Consumer Insights (Jul 2012 – May 2013) Brierley & Partners, Plano, Texas, USA

Addressed analytical issues facing major brand retail industry including Gamestop and Express apparel, help the clients build loyalty models to enhance the CRM program efficiency and optimize the marketing strategy.

Developing SQL and SAS codes to querying databases, perform comprehensive campaign analysis and automate the metric validation of the quarterly reporting processes.

Used database analysis to identify the key factors that drive predictive consumer behaviors so that to improve ROI. Sr. Statistician, Marketing Intelligence (Feb 2008 – Jul 2012) Javelin Marketing Group, Irving, Texas, USA

Provided statistical expertise to support various decisions making from the department of marketing intelligence.

Used SARIMAX technique to develop marketing-mix optimization integrated system for AT&T’s over 5 million dollars advertising program of wireless and wireline products. Helps AT&T optimally allocate market spending across different regions, different media vehicles, and different time period and develop meaning and actionable insights based on media/program decomposition results.

Built various experiment designs and help clients such as AT&T to optimize their marketing strategies.

Developed credit risk score model for banking client First Tennessee Bank, optimized the direct mail list and successfully reduced the chargeoff rate by 5%.

COMPUTER & PROGRAMMING SKILLS

Expertise in statistical applications/analytical tools such as SAS Enterprise Guide 4.3 (i.e., SAS/Base, SAS/STAT, SAS/ETS, SAS/SQL, SAS/GRAPH, SAS/AF, SAS/SCL, SAS/Macro), JMP 9.0, MS SQL Studio and Excel VBA.

Knowledge of Python 3.7 and packages such as Numpy, Pandas, Matplotlib, Scipy, seaborn, Scikit-Learn, and familiar with IDE such as PyCharm, Jupyter Notebook.

Strong PC skills in Microsoft suite of products and Unix/Linux shell tools such as Putty and Ubuntu. WORKING STYLE

Goal-oriented, highly motivated self-starter with a strong orientation to customer service.

Exceptional analytical and problem solving skills.

Confident, analytical in nature, fast learner and hard-working, strong team player.

Effective interpersonal and written/verbal communication skills. EDUCATION

Ph.D in Statistics, University of Louisiana at Lafayette, LA o Dissertation: ANOVA and MANOVA under Heteroscedasticity; Advisor: Dr. K. Krishnamoorthy

M.S. in Statistics, Sam Houston State University, TX

B.S. in Electrical Engineering, Xi'an University of Technology, China

Contact this candidate