Predictive Modeler/Statistician - Allstate (Menlo Park, CA)

POSITION AVAILABLE
ALLSTATE RESEARCH & PLANNING CENTER

The Allstate Research and Planning Center in Menlo Park, CA is seeking experienced predictive modelers and statisticians to help answer the next generation of research questions in the insurance industry. Our team uses the latest modeling techniques to build sophisticated customer pricing models and gain insight into the causes of insurance claims. Past projects include:

Using Geographic Variables to Improve Customer Pricing – An analysis of several thousand geographic variables (e.g. weather and climate information, purchasing patterns, population characteristics) to improve the accuracy of our pricing models. Project involved data visualization with SAS and MATLAB, large-scale variable selection techniques, and GLMs.

Identifying Fraudulent Theft Claims – Construction of a statistical model based on claim facts and policyholder information to predict fraudulent auto theft claims. Techniques included graphical data exploration, logistic regression, and decision trees.

We are looking for applicants with both exceptional modeling skills and the ability to communicate their findings clearly and concisely.

RESPONSIBILITIES
* Apply analytical expertise to solve real-world business issues.
* Develop research plans and collaborate with fellow analysts.
* Analyze and understand new data sources. Perform high-quality data preparation (data cleansing, feature extraction, merging data sources).
* Use SAS and/or MATLAB to perform analysis and build predictive models.
* Evaluate assumptions and stability of results on holdout data.
* Communicate effectively with business stakeholders through various stages of the project.
* Work with other teams to integrate findings/models into company operations.

SKILLS/EXPERIENCE REQUIRED
* Graduate-level degree with concentration in a quantitative discipline such as statistics, applied mathematics, or operations research.
* 2+ years of relevant work experience.
* Experience applying multivariate regression, data visualization, clustering and other data mining techniques to large-scale data sets.
* Programming experience with SAS and/or MATLAB
* Experience working in a UNIX environment. Perl and Python a plus.
* Proficiency with Microsoft Office (Excel, Word, PowerPoint).
* Team player but able to work independently with minimum direction.
* Strong writing and presentation skills.
* Must be eligible to work in the US on a full-time basis. Allstate does not sponsor individuals who are not already authorized to work in the United States.

Please send resume and cover letter to arpcjobs@allstate.com

Allstate is an equal opportunity employer.

This job ad has expired. Click here for active job ads.