As this book is completely dedicated to unsupervised algorithms, I've decided not to show a simple cluster analysis as a hello world! example, but rather a quite basic generative model. Let's assume that we are monitoring the number of trains that arrive at a subway station every hour because we need to ascertain the number of security agents required at the station. In particular, we're asked to have at least one agent per train and whenever there are fewer, we're going to pay a fine.
Moreover, it's easier to send a group at the beginning of every hour instead of controlling the agents one by one. As the problem is very simple, we also know that a good distribution is the Poisson one, parameterized with μ, which ...
