🎨 Custom Dataset
Table of Contents
You may not be interested in any of the pre-selected datasets, or may already be doing research/other work involving a dataset that is of particular interest to you. If that’s the case, you may be permitted to use a different dataset. You can take a look at our resources for finding datasets if there’s another topic you’re interested in. To request approval, send Suraj (rampure@umich.edu) an email with answers to the following questions:
- Why do you want to work with this dataset over the default options?
- What preliminary research questions do you plan to address with this dataset?
- Which features in the dataset do you consider key to your initial analysis and why?
- What feature do you plan to predict, and what features do you anticipate using in your predictive model?
- What is the size of the dataset (number of observations and features)?
- Is there missing data? If so, how do you plan on handling it?
- Please provide a link / attach the dataset, and include a screenshot of it already loaded into your notebook.
The deadline to request approval to use a different dataset is before the deadline for the checkpoint. You must email Suraj by Wednesday, March 26th. Once you email Suraj with the above, you should hear back within 48 hours with your approval or denial and reasoning. Note that submitting the checkpoint is not the same as requesting approval; the only way to request approval to use a different dataset is to email Suraj with answers to the questions above. After March 26th, if you haven’t requested approval, you must choose one of the default three options.