PEJMAN MAHBOUBI

!

***** ******** **. ******** *****, CA 913**-***-*** 0249 acgbws@r.postjobfree.com

Citizenship: US and IRAN

!

PROFESSIONAL EXPERIENCE

****-**** ****

HRL Laboratories (Contractor) – Malibu, CA

Research staff

Worked in a team to build an early warning system for the Boeing Company.

The data was a time series of maintenance messages (MMSG) that came from the air-

plane in real time with approximately 15000 different values. These messages were

occasionally accompanied with another message that demanded a repair (with appro-

ximately 10000 values). The value of the repair variable depended on a history of the

previous MMSGs. We built a system that predicted the coming repair message with an

80-85% accuracy.

My tasks were data preprocessing, data cleaning and feature selection.

Techniques: Machine learning, Predictive models, data visualization, …

Software: SVN, Gephi illustrator, Python’s Panda, SciPy and Matplotlib

Conducted a research project about human mobility patterns using twitter meta data. Users’

tweets that are posted from different locations reveal the users’ mobility patterns. We

used the tweet time and GPS to statistically analyze the temporal and spatial gaps

between consecutive tweets. We found that the temporal gap between any two

tweets with distinct GPS, follows a ‘Power Law’ distribution or Levy flight.

Techniques: Big Data, Statistical Analysis

Software: Hadoop, Pig and its “user defined functions”, Python, Java

Studied effects of the social media entries on products sales. For this project, we selected

every model made by three major car manufacturers and monitored 12 websites of the

corresponding dealerships. We were able to produce a time series of the sales number

for each car. We also analyzed the twitter data and produced a time series of the num-

ber of tweets for each car. Then we measured dependence between the two time se-

ries by computing their entropy correlation coefficients.

Techniques: Big Data, Web Scraping, Statistical Analysis

Software: Python’s Selenium, Hadoop, Pig, R, Matlab

PROFESSIONAL EXPERIENCE (ACADEMIC)

University of Utah, College of Arts and Sciences – Salt Lake City, UT 2012-May 2013

Postdoc

Collaborated with the probability team in research projects

Gave a scientific talk in a departmental seminar

Planned and taught courses in complex analysis and trigonometry

Department Of Mathematics, University of California, Los Angeles

Teaching Assistant and Graduate student

Completed PhD dissertation in probability

Title: “Existence and Regularity of Density for a Stochastic Heat Equation”

Led discussion sections for a class of 20-30 students

Courses: Probability, Calculus, Real and Complex Analysis, …

Summer research: Estimate of Sewall-Wrights density for small rate of mutation

EDUCATION

University of California, College of Arts & Sciences – Los Angeles, CA 2005-2012

PhD– Mathematics: probability, stochastic modeling and applied mathematics

University of Utah, College of Arts & Sciences – Salt Lake City, UT 2002-2005

B.A. – Mathematics, 2005

TECHNICAL SKILLS

Computer Languages: Python (Scikit-Learn, Pandas, Selenium, Network X), Java

Big data: Hadoop, Pig and udf, Machine Learning

Tools: SVN, Git, Gephi(dynamic and static illustrator), R and Matlab

!

PUBLICATION & PAPERS

“Analysis of the gradient of the solution to a stochastic heat equation via fractional

Brownian motion”, Submitted with M. Foondun and D. Khoshnevisan, 2014

!

!

Contact this candidate