My bachelor thesis: Exploring Performance of B-Tree Configurations Using Machine Learning Techniques

Structure

This project is organized into several main directories:

• code: This directory contains all the source code for the project. Data generating files are here aswell. It is further divided into subdirectories, each of which corresponds to a different part of the project. :
  • 01_preprocess: Contains notebooks for preprocessing data. See preprocess.ipynb and preprocess_ps.ipynb for more details.
  • 02_analyze: Contains code for analyzing the preprocessed data. Correlation and other exploration is done here.
  • 03_CNN: Unused. Initial experimatation for training a CNN using tensorflow. At the end not used for thesis.
  • 04_classifiers: Contains various classifiers and the XAI algorithms run on selected models.
  • 05_regressors: Contains various regressors and the XAI algorithms run on selected models.
  • 06_XAI: Unused. Initial experimentation of XAI algorithms was done here.
  • 07_leaf_size: Unused. Data exploration of the old pattern recognition dataset was done here.
  • 08_pattern_recognition: Contains the pattern recognition algorithms. The only relevant file is apriori_small.ipynb
• output: This directory is used to store output files generated by code/generate_data.py.
• latex: Contains the LaTeX code for the thesis. See the README.md in this directory for more details and references to original author.
• Bachelor_Thesis_final.pdf: Is the final thesis in pdf format.

Setup

Make sure to have python and poetry installed:

poetry install

poetry run python code/generate_data.py

I use VS Code for jupyter notebooks.

Notes

The btree-binaries, the urls-data and outputs are hidden, since they are too big.

Some images are not published outside the pdf, due to copyright concerns.