Getting started with Kaggle

Kaggle image

Table of contents

  • What is Kaggle ?
  • Why Kaggle ?
  • Your first ML model on Kaggle

Kaggle is the world’s largest data science community with powerful tools and resources to help you achieve your data science goals. —

Basically Kaggle is a platform dedicated for datascience where you can learn, share and discover a lot of ressources. This platform offers :

  • Data : The users can share and use data on the platform, those datasets are sometimes provided with a description and some other details to comprehend them easily.
  • Notebooks : If you are not familiar with this term, a notebook is a document that can contain organized text (better than comments) and run lines of code. Users can share and access notebooks provided by other users, so that is knowledge sharing!
  • Competitions : The platform allows its users to participate in datascience competitions with money as a prize (mostly dozens of thousand of dollars minimum).
  • Courses & Communities : The platform allows its users to learn datascience or machine learning and stuff by using the Courses ressources and users can share/exchange on the platform too with Communities.

Kaggle is free and most of its ressources are made by users who are either professionals or students in the sector. So you don’t have to be afraid of learning wrong stuff because the community is large enough to help each other.

Now before getting started you will need some tools :

  • a brain
  • a computer

Just kidding! So let’s get into it.

  1. Go to the Compete page
Compete page

2. Click on this competition : Titanic — Machine Learning from Disaster

Titanic — Machine Learning from Disaster competition for knowledge

3. Click on Notebooks/New Notebook

Kaggle — Notebooks
Kaggle — New Notebook

Just keep those settings.

4. Write some code

Kaggle — Notebooks | Predictions based on ‘Sex’ column

5. Press on Run All

6. Refresh the directory, press the More actions button and Download

Kaggle — Refresh | More actions/Download

Then the submission.csv file will be downloaded on your computer.

7. Go the the competition page and click on Submit predictions/Make Submission

Kaggle — Competition page
Kaggle — Make Submission

Upload your submission.csv file on Step 1 and write a small description for your submission just in case you will make many others.

8. Check your score/leaderboard

Kaggle — Score

Here I predicted 76.5% right for the survivors of Titanic! Now as a data-scientist you will have to modify this model to make a better model for better predictions.

Hope that you learnt something, here is my Kaggle profile if you want to check some of my work .

Analytics Vidhya

Analytics Vidhya is a community of Analytics and Data…