Projects & Reports
Bayesian Statistics project
This project defines a method to fit a regression model for given data using regularization penalties. The results from this project can be used effectively with sparse, high-dimensionality data. Briefly, the project comprised of:
-
Find first-stage coefficients by using OLS/Ridge
-
Using the scale of first-stage coefficients as weights, find Lasso coeffecients using a deterministic algorithm
-
Validate optimal coefficients using Leave-One-Out-Cross-Validation
Maintenance downtime reduction project
I first collected down-time data from the shop-floor. Briefly, the project comprised of:
-
Developing a protocol for digital entry of maintenance activity
-
Identifying bottleneck processes through downtime analyses
-
Building predictive models for critical spares
-
Standardizing spares
-
Implementing "Kaizens" to improve critical processes
Regression Analysis project
A term project in a graduate class on Regression using a method typical of such an analysis - parameter estimation, variable selection, model building, data validation.
Design of Experiments project
Designed and performed a factorial experiment with blocking to determine significant factors affecting cell-phone battery life.
Stochastic Simulation project
A term project in a graduate class on simulation of Stochastic events. We conducted the project with Einstein Bros, a bagel shop on campus. Simulation was done in ARENA. Briefly, the project comprised of:
-
Collecting real-time queueing data and modeling an input distribution
-
Simulating several scenarios to determine key metrics of the queue