🚀 Tasks

As always, the data for this task is on moodle. Follow the instructions in Submission via GitHub Pages to create a new repo in the existing organization.

Your task is to improve upon the simple model we have built in the chapter Minimal example. You can do this by:

1. Feature Engineering: Create additional features that you think might help to distinguish between different transportation modes. For example, you could consider turning angles, stop durations or extending observation windows
2. Model Selection: Experiment with different models (e.g., random forests, gradient boosting, neural networks) and hyperparameters to improve the predictive performance.
3. Per point prediction: Instead of predicting the transportation mode on a per segment basis, try to predict the transportation mode for each point. This is a more challenging task but can provide more detailed insights into the movement behavior.

Using one or more of these approaches, predict the transportation mode for the validation dataset.