Exercise

Handwritten digits dataset

The handwritten digits dataset has around 160 records for each digit ranging from 0 to 9. Each record is formed by 256 columns that takes the value 0 or 1, representing an image of 16x16 pixels.

Later in the course, you will use this data to identify the clusters that can explain the digits using flexmix. Now, the aim is to get familiar with datasets that can be modeled with mixture models using Bernoulli distribution.

The object data has the observations of only two digits from the original dataset.

Instructions

100 XP
  • Use dim() to check the dimension of the data frame.
  • Use glimpse() to check the data. This is a handy function to check every column, especially when there are many.
  • Use the function show_digit() to parse the rows 50 and 100 into an image.