Data Scientist

Xiao Xu

****.**@*******.*** 630-***-**** (US) +86-150******** (China)

Resourceful data analysis professional with extensive experience in data mining, predictive analytic modeling, and statistical analysis. Have developed a strong background for bringing innovative statistical insights to demographic and psychographic profiling and predictive data modeling. Particularly adept at customer profiling, sales measurement design to guide decision-making and meet customer satisfaction. Strong communication, presentation, collaboration, and teamwork capabilities used to drive consensus among diverse stakeholders and bring projects to fruition.

PROFESSIONAL EXPERIENCE

2015-present Division Leader/ Senior Technical Specialist Sinopec Group, Qingdao, China

Building and leading rotating machinery safety analysis technical devision at Sinopec Safety Engineering Research Institute to develop technologies for rotating machinery safety operation management across Sinopec petrochemical plants and refineries

Establishing group-level rotating machinery safety management platform

Developing rotating machinery safety management platform based on AI and data mining technologies. To provide large-scale real-time health monitoring on rotating machinery and processes across Sinopec petrochemical plants and refineries

Developing models of fault analyses and fingerprint database

2010-2014Project Manager / Senior Statistician, InterPublic Group, Downers Grove. IL, USA

Developed Enhanced Profiling process to provide scientific base for pre-print advertisement campaign planning

Established a process by applying a combined statistical and AI modeling approaches in SPSS to quantitatively provide consumer target profile and to optimize campaign planning/scheduling for maximizing customer targeting and sales lift

Providing clients with business insights using data mining and modeling techniques using SPSS tools, Matlab toolboxes and R packages, as well as R programming

Massive client sales data acquisition, cleasing and relationship discovery between consumer demographic/psychographic measures and geographic sales to identify sales pattern and comsumer targeting profile

Enhanced profile modeling for more than 10 national retail chains to provide insights for internal and external teams to guide optimal media placement planning using CART, regression, tree analysis, neural networks, clustering approaches

Providing consumer insights for clients to identify potential consumer targets

Designing control & test plans for clients to measure media effectiveness

Developing methodologies and tools suitable for media audience targeting

Developed a VB tool in MSOffice to calculate audience/target numbers and density by applying a developed a calculation formula on demographics to reveal media audience quantity in certain geographic boundaries to provide heat map visualization for clients on media placement guidance

Providing clients guidance for advertisement budgeting allocation

Providing national/regional demographic and psychographic insights to clients to reallocate geographic-based print budgeting to maximize ROI in advertisement

2007-2010 Technology Leader, R&LT, BP North America, Naperville, IL, USA

Developing and applying data mining technologies suitable for rotating equipment health/performance monitoring and for process analysis in BP refining business

Non-parametric modeling for refining process fault analysis using statistical and AI approaches

Developping condition monitoring technologies for refining process health monitoring, process prognostics/diagniostics and process optimization using Matlab with different data driven methods (kernel regression / linear and nonlinear regression methods, PCA, PLS, etc) and performance-improved fault alerting mechanism

Inferential modeling as virtual analyzer

Data-based column process modeling for product yields estimate

Managing projects and developing data mining & predictive analytic technologies in equipment health monitoring and refining process in BP’s refineries

Responsible for developing analytic models for refining equipment to identify potential faults in saving downtime with savings at about 16%

Tibco Spotfile visualization for comparing movements of various machinery operation parameters to identify potential issues

2004-2007 Project Manager / Sr. Research Technologist / Sr. Application Engineer, SmartSignal Corporation, Lisle, IL, USA

Managing and Developping machinery condition monitoring projects in power and transportation industries

Developping ad-hoc tools for predictive condition monitoring

2000-2004Research technologist, SmartSignal Corporation, Lisle, USA

Developed next generation of SmartSignal’s equipment condition monitoring software product for applications in industrial process health monitoring, process prognostics/diagniostics and process optimization

Studying different data driven methods (kernel regression / linear and nonlinear regression methods, PCA, PLS, etc)

Successfully applied in avaition, power and other transporatation industries

Developed better behaved fault alerting engine for software product

EDUCATION

1995-2000 Doctor of Philosophy in Nuclear Engineering - speciality in artificial intelligence / data mining and modeling / pattern recognition University of Tennessee, USA

1992-1995 Master of Science in Nuclear & Energy Engineering University of Arizona, USA

PATENTS

Possessing four US patents in AI, data mining and pattern recognition applications

Contact this candidate