Kaggle competitions
It’s good to know that this project will be continuously updated whenever I participate in another competition on Kaggle. If you’d like to stay up-to-date on my Kaggle adventures, the best way is to connect on Kaggle. You’ll then automatically receive an email with an update to a competition.
Titanic
November 2024
Predict the survivors on the Titanic
The goal of this competition on Kaggle is, to use Machine Learning to determine which people survived the Titanic disaster and who did not. It is considered to be a good first competition to get your hands dirty on Machine Learning.
As a model I used the RandomForestClassifier, which is a decision tree algorithm. 1 Decision trees operate using an algorithmic approach that splits the dataset into individual data points based on various criteria. These splits are performed using different variables or different characteristics of the dataset..
As features I used the following data:
- In which class a passenger was
- Male/female
- Whether and how many siblings/spouses were on board
- Whether and how many parents/children were on board
My goal was more to understand Machine Learning and scikit-learn than to actually compete.
If you want to know more about this project, you can look at my code on Github or take a look at the page on Kaggle.
Spaceship Titanic
March 2025
Predict which passengers will be transported to an alternate dimension.
Spaceship Titanic is a bit of a sequel to Titanic, the previous competition where you had to predict the number of survivors. In Spaceship Titanic, the passengers are transported to an alternate universe/dimension.
Here my goal was to work with Machine Learning Pipelines (ML Pipelines) and to understand the difference between RandomForest and Gradient boosting.
