Homework: Gather a dataset.
In the following few lessons, we'll cover various techniques to help you explore your data and reveal its characteristics.
Learning to explore a dataset works best if you are interested in the data itself.
Your task until our meeting next week is to find or create a dataset on any topic that you find interesting.
For now, do not care too much about the nature or modality of the dataset, but make sure that:
- It contains numerical (meta-) data
- It has more than 50 entries
- More than two values/features should describe each record
If interested in analyzing text-based data, you can leverage your BagOfWords encoder to create word frequencies that numerically describe the texts.
Furthermore, prepare a short (3-5 min) presentation describing your dataset, its features, why you chose this particular dataset, and what question do you like to investigate using the dataset.
If you'd like to, feel free to prepare slides for that, but you are not required to do that.