Engineering Consulting Projects
This section contains a non-exhaustive list of Statistics consulting projects I worked on which subject matter relates to the field of engineering. A link to the final project outcome is provided when the client allows for it.
Analysis of Crew Members Performance in NASA BASALT Mars Simulations.
August 2016 -- May 2019 · Industrial Engineering R | SAS
August 2016 -- May 2019 · Industrial Engineering R | SAS
- Analysis of astraunots live physiological monitoring data and performance obtained during extravehicular activity (EVA) in BASALT Mars simulations .
- Built numerous non-parametric models for evaluating crew health and performance and comparing crew physiological data distributions.
- Assisted client with data preprocessing, code writting for models, results interpratation and report.
Quantifying Insurance Providers Network Graphs Effect on Patient Data
January 2017 -- May 2019 · Industrial Engineering R | SAS | SQL
January 2017 -- May 2019 · Industrial Engineering R | SAS | SQL
- Analysis of a 5.5 years longitudinal health claim data trough a provider/patient social network analysis.
- Investigated the effect of providers network graphs measures (degree, centrality and betweeness) on very sparse patient level count data.
- Modelled the provider network graph effects using a zero inflated Poisson model.
- Established the covariance structure of the various models through a Generalized Estimating Equation (GEE) method to caputure the depentence between patients in the same network clusters.
- Used boostrap method to compute confidence region of various statistics.
- Assisted client with code writting for models and results interpretation and reporting.
Disease Status Prediction and Growth Trend of Abdominal Aortic Aneurysms
August 2017 -- May 2018 · Biomedical Engineering R | SAS
August 2017 -- May 2018 · Biomedical Engineering R | SAS
- Used high frequency ultrasound aneurysm data to develop prediction models of both aneurysm formation and growth trend.
- Used a quadratic discriminant analysis and logistic regression to build two statistical models to predict disease status.
- Validated model performance through leave one out cross validation prediction accuracy, ROC curve, specifcity and sensitivity analysis.
Engineering Teamwork Ratings Quality Assessment
August 2016 -- May 2018 · Engineering Education R | SAS
August 2016 -- May 2018 · Engineering Education R | SAS
- Developed metrics and statistical methods for accessing the quality of ratings data generated by the CATME system.
- Built tools for automating data cleaning and analysis in SAS and R and created instruction manuals to help CATME system users run analyses independently.
- Run live demo of new method for CATME users from different countries.
- New tools are to be integrated into the CATME system which is currently deployed to more than 1 million users from 80 countries.