Summary

Excellent Quantitative analyst in Risk management and marketing with 10 years experience, especially in advanced approach methods in credit risk and operational. Especially in advanced analytics, predictive modeling, with 15 years experience in finance, retail, healthcare, marketing and insurance and clinical trial.

High skilled statistical programmer with more than 15 years in SAS, SAS Macro, SQL, R, SPSS and other statistical software.

Working Experience:

Confidential

• Cancellation risk models of EMM Customers by Tenure Groups
• LTV analysis by using survival analysis
• Profitability analysis of Email Campaigns based on the LTV
• Trialer conversion models by trail time
• Cancellation analysis of ToolKit customers and Single Plat Form Customers
• Cancellation Risk Model of ToolKit Customers
• Projection analysis of Single Plat Form Customers by using time series and Simulation method

Confidential

• The projects include
• Risk management of financial unexpected loss,
• Predictive models of fraud detection of medical and pharmacy claims
• Predictive models on bad clinical outcome of chronic diseases in medical management practices, these bad clinical outcomes are frailty, rehospitalization, and etc.
• Financial impact analysis on the new PPS from CMS.
• The predictive model of 30-day rehospitalization has been adopted by AHCA and CMS as new Quality measure for SNF skilled Nursing Facilities .
• Recent publication related to the current work:
• John Gao etc Financial Impact Analysis of New RUG-IV for Post Acute Medicare Provides using Monte Carlo Simulation , in NESUG Nov 2012
• John Gao etc Eigen Knots In Spline Logistic Regression Method in NESUG Sept 2011

Confidential

• Developed predictive default risk model and prepayment model of private student loan at origination
• Developed refresh default risk and prepayment at the private student loan entering repayment
• Forecasting default risk at portfolio level by using time series
• Simulation analysis of default risk for short term and long term at trust and segment level TIER
• Estimate the recovery rate for short term and long term at trust, segment and overall portfolio of asset
• Cash flow analysis at trust level, segment and overall portfolio of asset, in which there are modeling analysis of PD, LGD and EAD
• Estimate the prepayment rate for short term and long term at different level
• Evaluation of overall NPV of asset life time
• Pricing analysis of the segments based on the cash flow analysis result.
• Developed predictive model of Fraud, non-payment default risk in loan application
• Several research results have been adapted by Moody, S P as a standard analytical method in the private student loan industry.
• John Gao et al. March 2008 predictive default risk model of private Student loan NBES 2008 conference meeting accepted for presentation .

Confidential

• Invitation generic consumer bankruptcy risk model by using non credit bureau information
• Predictive consumer bankruptcy risk model by using credit bureau information
• Predict the market value of real estate properties by using the information from public record such as tax , property transaction records and etc .
• Estimate the current LTV the ratio of loan to market value for property owners by using advanced statistical method, such as mixed model.5
• Develop Bias Reduction models by using credit information from 3 credit bureau companies.
• Also, I have developed response models by using national data base to support marketing campaign of clients

Confidential

• Response model of paper buy
• Response model of ink buy
• Credit card acquisition model
• Response model of Mail/Ship
• Early warning model of customer purchase pattern broken
• Marketing Efficiency Analysis of marketing multiple campaign actions
• DM Coupon Response/Self Motivation model of Small Business Customers

Confidential

• Develop quantitative model or analytical models for example, advanced stochastic method in operational risk, cash flow analysis for deposit banking customers, predictive model of customer attrition, cluster analysis of transaction pattern of customers, cross sell for marketing, business operation, and risk management etc using customer data collected in data warehouse, data mining and advanced statistical methods.
• Operational Risk: Leader of Quantitative Analysis Team in BASEL II team in Fleet. The projects include:
• Developed PLDA predictive loss distribution approach for predicting the probability of future loss and has been successfully implemented in capital allocation in FleetBoston.
• Trend analysis of operational loss in retail, small business
• Conduct monthly analysis report of operational loss events in retail and small business
• Implemented EVT model, AMA and PLDA methods for the capital allocation.
• Evaluated the efficiency of operational risk programs, such as fraud detecting, internal management and etc and internal audition.
• Estimated the benefit of loss mitigation
• Studied the strategy to reduce check transition fraud, including predictive model, software implementation and etc.
• Developed a new spline regression method. A comparison study shows that the new method can conduct better result than other data mining methods such as CART, MARS and Neural Network in prediction, stability and accuracy.