Present to the class some cutting-edge research on the brain and how it works
Train a neural network to recognize handwritten
Sütterlin (ask your grandparents what this is
)
For some large dataset, train a neural network and do some analysis on model parameters such as number of hidden nodes, learning rate, momentum, etc.
Perform some deep network analysis on the
Pleonast user graph (which I will provide)
Do a Milgram-esque social experiment testing the Small Worlds hypothesis (perhaps by email?)
Determine some sort of social graph for HftL students
Do some network analysis on various network topologies such as random, “connected caveman”, k-lattice, scale-free, etc.
Develop a business plan to take advantage of a long tail