User:Dakota.Sullivan

From REU@MU
Revision as of 18:08, 19 August 2015 by Dakota.Sullivan (Talk | contribs)

Jump to: navigation, search

Personal Info

Dakota Sullivan is a Computational Mathematics and Social Welfare and Justice double major at Marquette University.

Research Topic

Using statistical surrogates to estimate the behavior of storm surge during a hurricane or tropical storm. See here for more details.

Milestones

Weekly milestones can be found on our main project page, located here.

Daily Log

Week 1 (6/1 - 6/5)

Monday, June 1

  • Attended introductory session; learned about REU program
  • Met with Dr. Spiller and research group members
  • Attended Systems lab orientation with Dr. Dennis Brylow
  • Took pre-REU NSF survey
  • Began introductory reading on Gaussian Stochastic Process

Tuesday, June 2

  • Toured campus
  • Filed paperwork with HR
  • Attended library resources orientation with Heather James
  • Met with Dr. Spiller

Wednesday, June 3

  • Installed Python
  • Met with Dr. Spiller to review Gaussian Stochastic Process
  • Met with Dr. Brylow to learn about Marquette email and time keeping
  • Tested optimization routines in Python

Thursday, June 4

  • Determined routine for plotting in Python
  • Attended talk by Dr. Factor
  • Met with Dr. Spiller
  • Determined research goals and project timeline
  • Continued developing optimization routine and plotting algorithms

Friday, June 5

  • Met with research group to finalize optimization algorithm
  • Reviewed objectives of project
  • Documented Goals and Milestones
  • Downloaded LaTex

Week 2 (6/5 - 6/12)

Monday, June 8

  • Met with Dr. Spiller
  • Created methods to begin modeling maximum likelihood equation
  • Developed a plot of our approximation in Python

Tuesday, June 9

  • Created maximum likelihood function
  • Worked through difficulties with singularity in our matrices
  • Plotted approximation of theta optimization

Wednesday, June 10

  • Corrected errors in the maximum likelihood function
  • Determined optimal theta value
  • Modified maximum likelihood equation to optimize multiple variables

Thursday, June 11

  • Developed variance function
  • Listened to sample presentation by Dr. Factor
  • Met with Dr. Spiller, updated her on our progress and discussed paper by Westerink et al.

Friday, June 12

  • Modified code to account for all approximated variables
  • Tested our approximation methods for other functions
  • Began reading Westerink et al.

Week 3

Monday, June 15

  • Read through research paper (Westerink et al.)

Tuesday, June 16

  • Met with Dr. Spiller
  • Introduced to additional notes on GaSP
  • Attended talk by Dr. Brylow

Wednesday, June 17

  • Compared previous and new maximum likelihood equations
  • Began developing a least squares model for our approximation
  • Met with Dr. Spiller to discuss new maximum likelihood equation

Thursday, June 18

  • Completed Responsible Conduct of Research (RCR) training
  • Began implementing new maximum likelihood equation
  • Integrated least squares model into maximum into approximation

Friday, June 19

  • Finalized implementation of least squares model
  • Tested approximation with multiple functions and inputs
  • Added reference prior to maximum likelihood equation
  • Met with Dr. Spiller
  • Watched RCR training videos

Week 4

Monday, June 22

  • Modified approximation function to allow multiple dimensions
  • Developed algorithm to import test data
  • Updated personal wiki page

Tuesday, June 23

  • Tested import algorithm
  • Met with Dr. Spiller
  • Searched for errors in code
  • Modified m(x) and v(x)

Wednesday, June 24

  • Began optimizing theta for multiple dimensions
  • Created algorithm for n-1 linear regressions

Thursday, June 25

  • Met with Dr. Spiller
  • Attended talk by Dr. Spiller
  • Modified Linear Regression
  • Plotted new approximation

Friday, June 26

  • Tested various functions with approximation
  • Cleaned/commented code
  • Began optimization for n dimensions

Week 5

Monday, June 29

  • Tested n-dimensional case with three parameters
  • Outlined presentation

Tuesday, June 30

  • Met with Dr. Spiller
  • Worked on presentation

Wednesday, July 1

  • Worked on presentation
  • Read “A Framework for Validation of Computer Models”

Thursday, July 2

  • Finalized presentation
  • Attended presentations

Week 6

Monday, July 6

  • Met with Dr. Spiller
  • Read through Step 5 of “A Framework for Validation of Computer Models”
  • Identified unknowns in process

Tuesday, July 7

  • Planned out MCMC example solution
  • Programmed MCMC solution

Wednesday, July 8

  • Worked on MCMC Weibull Example
  • Read through “Posterior Models and MCMC”
  • Worked on Beetle Mortality Example

Thursday, July 9

  • Met with Dr. Spiller
  • Work Lunch
  • Altered Code for Weibull Example
  • Plotted posterior function against MCMC histogram

Friday, July 10

  • Attempted to use GaSP Approximation to determine bias function
  • Obtained estimate for bias
  • Finished Weibull Example

Week 7

Monday, July 13

  • Updated Wiki
  • Completed RCR quizzes
  • Identified problems in current code
  • Met with Dr. Spiller
  • Attempted to remove noise from function approximation

Tuesday, July 14

  • Completed thinning algorithm
  • Verified accuracy of MCMC method
  • Searched for errors within bias function
  • Met with Dr. Spiller

Wednesday, July 15

  • Searched for cause of high variance
  • Adjusted minpoint method
  • Printed all variance values
  • Plotted optimization function
  • Verified that problem exists within the maxprior function

Thursday, July 16

  • Met with Dr. Spiller
  • Plotted approximation without prior
  • Tested x matrix as a vector of ones
  • Work lunch and presentation

Friday, July 16

  • Determined correct alpha values

Week 8

Monday, July 20

  • Met with Dr. Spiller
  • Tested alpha values for approximation and bias functions
  • Started outline for poster

Tuesday, July 21

  • Reviewed Bayari paper
  • Worked on presentation

Wednesday, July 22

  • Met with Dr. Spiller
  • Fixed maximum likelihood equation
  • Created MCMC algorithm

Thursday, July 23

  • Worked on latin hypercube algorithm
  • Work lunch and talk
  • Modified MCMC algorithm to work with two dimensions

Friday, July 24

  • Modified MCMC algorithm for n dimensions
  • Worked on presentation

Week 9

Monday, July 27

  • Worked on presentation

Tuesday, July 28

  • Worked on presentation
  • Outlined research paper

Wednesday, July 29

  • Finalized presentation
  • Worked on abstract

Thursday, July 30

  • Worked on research paper
  • Submitted abstract
  • Met with Dr. Spiller
  • Work lunch
  • Continued to work on research paper

Friday, July 31

  • Worked on presentation
  • Worked on research paper
  • Tested MCMC algorithm

Week 10

Monday, August 3

  • Met with Dr. Spiller
  • Worked on presentation
  • Tested MCMC algorithm

Tuesday, August 4

  • Met with Dr. Spiller
  • Worked on presentation
  • Poster presentations

Wednesday, August 5

  • Finalized presentation
  • Presentations
  • Work lunch

Thursday, August 6

  • Presentations
  • Worked on MCMC algorithm

Friday, August 7

  • Post-REU Survey
  • Final meeting with Dr. Spiller