/bsc_thesis

Bachelors thesis from Universität Leipzig. Topic is machine learning analysis of predicting pedestal plasma density using the JET pedestal database. The data can be accessed according to the data access rules within EUROfusion.

Primary LanguageJupyter Notebook

B.Sc. Thesis Universität Leipzig 2021

Topic: Analysis of machine learning towards predicting plasma density using the JET pedestal database.

Advisor(s): Frank Cichos & Mathias Groth & Aaro Järvinen

Outline of repository

  • main thesis document, thesis.pdf, and presentation showing main results pres.pdf are found in the home dir.
  • src/ holds the code, where experiment scripts are found locally, and reused methods like dataprocessing are found in src/codebase/
  • /src/out/ holds the data used for plotting found in the thesis
  • doc/ is for latex
  • etc/ is for brainstorming

Experiments output either a .pickle or .txt file, but the specifics are found in each script. Most experiments use arguement parsing so you can throw everything in a bash script and send it to your favourite cluster.