Skip to content

Files

Latest commit

57fe26e · Sep 24, 2020

History

History

Training Better Models Using Automated Machine Learning

Here's what you need in order to run this notebook:

  • Clone this repository somewhere nice
  • Create a Kaggle account, join the Titanic competition competition, download its data files to the data folder
  • Make sure you have an Azure account, in it create a resource group and a Machine Learning Service Workspace - you'll need those for running the automated machine learning experiment
  • Download config.json from your Azure ML Workspace to this folder.
  • Install either Miniconda or Anaconda. The Python 3.7 version 🐍
  • Run the following commands inside this directory
conda env create -f env.yml -n automl_on_titanic
conda activate automl_on_titanic
ipython kernel install --user --name=automl_on_titanic

jupyter notebook
  • If everything runs fine and at the end you can open the auto-ml-and-kaggle.ipynb notebook then you're good to go.
  • After you're finished, you can run conda remove --name automl_on_titanic --all to remove the environment you had created.