Banner Image

Skills

  • Data Management
  • Data Mining
  • Machine Learning
  • Modeling
  • R
  • Statistical Analysis
  • Surveys

Services

  • Professional Statistician

    $75/hr Starting at $150 Ongoing

    Dedicated Resource

    I can assist you with all of your statistical needs including: Design and Analysis of Experiments Survey Design Segmentation Classification and Data Mining Bayesian Hierarchical Modeling Causal Modeling...

    Data ManagementData MiningMachine LearningModelingR

About

PhD Statistician specializing in advanced research requiring customized research design, statistical analyses, and consultation.

Mark has been a Senior Manager with the Statistical Analysis group at comScore, a Senior Statistician in the Data Sciences group at Compete (a MillwardBrown Digital company), and a Director of Marketing Sciences with in4mation insights. Prior to moving to industry, Mark was a Lecturer in Statistics at Harvard University and an Assistant Professor and Statistical Computing Scientist in the Statistics department of Ohio State University.

Most recently, I have worked in the field market research developing methods for the analysis of internet clickstream data. I lead the development of MBD’s Consumer Behavioral Impact product, which efficiently estimates the lift
of Internet advertising on Key Performance Indicators (KPIs) while accounting for selection bias. I also lead the development their Path to Purchase product, which mines Internet clickstream data from a large multisource panel to cluster how people use the Internet to investigate products prior to purchase.

While with in4mation insights, I performed custom analyses for many clients, including implementing a store location model for a national supermarket chain based store characteristics, demographics and competition, a pricing model for a major online retailer based on sales history and competitor pricing, and a portfolio churn predictor for a large overseas communications company.

Prior to moving into market research, my research interests in the past have included statistical genetics, focusing on problems in gene mapping, space-time hierarchical modeling, and statistical computing, dealing mainly with Monte Carlo methods. Past projects have dealt with modeling daily ozone data in the Lake Michigan region, class discovery and classification of tumor samples using gene expression data, the effects of interference in haplotype reconstruction, examining changes in military defensive strategies through point process models, and spatial-temporal nonlinear filtering in Command and Control.