Task: Pet Registration and Trends
Background:
Under Domestic Animals Act 1994, all domestic cats and dogs in Victoria over three months old must be registered with the local council. The database available at
(https://www.data.gov.au/dataset/dandenong-registered-cats) and (https:// www.data.gov.au/dataset/dandenong-registered-dogs)
contains information about the suburb and postal code of each registered animal, the primary breed, primary colour, its month of birth, and whether it has been desexed.
There are over 200 dog and 50 cat breeds listed (though some are crosses) and a comparable number of colours, though, again, some are mixes. You may need to think about how to handle such a large number of levels of a factor. You may also want to look up the demographics of the suburbs in question over time.
1
Requirements:
Download the latest version of the datasets from the websites above.
Explore the relationship between the suburb (and its demographics) and the popularity of the breeds and colours of cats and dogs registered there, as well as their popularity over time (as measured by year and month of birth) using visualisation, numerical summaries, and models, and present your results.
Is there evidence of changes in relative popularity of cats and dogs between petsborn up to 2014 and those born later? Does the change between these two periods differ by postcode?
Propose two different models to predict whether an animal will be desexed basedon other information available (species, location, when born, etc.).
Present details of data pre-processing. This could include some merging of categories, representing mixes differently, and other techniques.
Discuss the strengths and weaknesses of proposed models, and present your preferred model.
Present detailed numeric and graphical performance measures of your classificationresults.
