Resume

Sign in

Sr Data Scientist, PhD, Data detective and machine learning expert

Location:
San Francisco, CA
Salary:
160000
Posted:
April 22, 2016

Contact this candidate

Resume:

Ragnar Lesch, PhD.

** ******** ***, **** ******, CA 94941 650-***-**** acugkx@r.postjobfree.com

SENIOR DATA SCIENTIST

A Senior Data Scientist credited with leveraging expertise in mathematics and applied statistics with programming skills to create predictive models and tools that increase business opportunities and profitability. Passionate about data and driven to uncover new insights in large sets of structured and unstructured data paired with a willingness to explore new territories to solve complex business problems.

Key Technical Skills

Tools:

SAS, SQL (Netezza, SQL Server, Teradata, MySQL), KNIME, Matlab, Python, C/C++, Java, Pearl

Modeling concepts:

Machine Learning, Time series analysis, Clustering, Generalized Linear and Additive Models, Nonlinear Regression, Classification, Neural Networks, Decision trees, Text mining

Big Data Tools:

Hadoop, MapReduce, HBase, Hive, Pig, Spark, Splunk

Special expertise:

Marketing Mix Modeling and Optimization, AB Test Design and Evaluation, Customer Targeting and Segmentation, Insurance pricing, Fraud Detection Modeling

Professional Experience

UNIVERSAL MCCANN (The Interpublic Group), San Francisco, SF 9/2010 – 6/2015

Advanced Analytics Partner

Led analysis of advertising impact on customer behavior by creating marketing mix models, and by designing and evaluating A/B experiments for various marketing activities; used models and tests to increase marketing effectiveness and overall profitability for national advertising campaigns

Designed a data-warehouse framework to manage transaction-level ad exposure and web browsing behavior data

Developed analytics reports based on those to deliver weekly updates about campaign results and created customized reports to support new business pitches

ARA CAPITAL, LLC., Berkeley, CA 9/2009 – 8/2010

Co-Founder & Director of Analytics

Created and implemented quantitative strategies for a market neutral U.S. equities hedge fund

Developed research platform for model design and trading simulation, built trade execution engine for signal generation and order management (Matlab, Java, SQL) using real-time and historical tick data

ALLSTATE RESEARCH & PLANNING CENTER, Menlo Park, CA 4/2007 – 9/2009

Predictive Modeler/Manager

Developed predictive models for pricing and risk indication while leading a team of Predictive Modelers to produce accurate and highly effective solutions to achieve business goals

Built a claim fraud model to cut costs on losses due to fraud, consulted deployment team for implementation

VALEN TECHNOLOGIES, Denver, CO 7/2005 – 3/2007

Senior Predictive Modeler

Designed and created predictive models for P&C insurance clients for loss prevention, risk evaluation and pricing

Implemented key algorithms in proprietary modeling tool for company-wide use.

Consulted clients in project management and business integration of predictive models with exceptional customer service and attention to detail

Professional Experience (Cont’d.)

KOSMEDIX, INC., San Francisco, CA 2/2003 – 6/2005

Co-Founder & Director of Technology

Managed as one of two founders all day-to-day activities of a small startup (IT, HR, accounting, web presence, Internet sales infrastructure)

Developed database warehouse for quantitative marketing of consumer goods across various media channels

Performed ROI analysis for online and offline campaigns and improved customer targeting by creating more accurate customer profiles

QUANTLAB FINANCIAL, Houston, TX 11/1999 – 2/2003

Quantitative Research Scientist

Re-designed and expanded procedures for derivatives-based predictive modeling in an R&D company that designs, develops and applies advanced modeling technology for the financial markets operating a $100 million hedge fund. Held full responsibility for data ETL, cleaning, validation, and signal generation,

Created a suite of C++ classes for daily and intra-day equity and option data (incl. the calculation of the implied volatility and other key derivative statistics).

Fine-tuned existing basic strategies, enabled a more reliable and successful predictive signal used in trading, and improved the overall quality of option signals.

Education & Training

Doctor of Philosophy Neural Computing in Finance Aston University Birmingham, UK

Master of Science Computer Science & Psychology University of Erlangen-Nuremberg Germany

Publications

Lesch, R. H., Caille, Y. and Lowe, D.: "Component Analysis in Financial Time Series", Proceedings of the IEEE 1999 Conference on Computational Intelligence for Financial Engineering (CIFEr '99)

Lesch, R. H. and Lowe, D.: "Towards a Framework for Combining Stochastic and Deterministic Descriptions of Nonstationary Financial Time Series", Proceedings of the 1998 IEEE Signal Processing Society Workshop: "Neural Networks for Signal Processing”

Memberships

International Institute of Forecasters (IIF)

Association for Computing Machinery (ACM)



Contact this candidate