You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Feature selection is a key step in machine learning as it boosts computational efficiency and predictive performance by keeping only the most relevant predictors.
Beyond the popular supervised feature selection classes like filter and wrapper methods, global search methods like simulated annealing are also powerful techniques at our disposal.
In this project, we delve into the theory and application of simulated annealing for feature selection.
Project Structure
data: Titanic dataset (raw and processed)
images: Set of images and visualizations used to demonstrate algorithm
notebooks: Jupyter notebooks for the different steps of the project i.e. data pre-processing, baseline modeling, and running of feature selection with simulated annealing algorithm
results: CSV files of the output from algorithm runs
src: Python scripts for simulated annealing algorithm for feature selection
main.py: Main script containing algorithm. In CLI, cd into src folder, then execute python main.py
utils.py: Utils script containing ML model function (i.e. random forest classifier)