Title: Zero-altered Lognormal Regression

Supervisors: Thomas Yee and Ian Tuck 

Prerequisites: STATS 310, STATS 330 and STATS 380

This project in statistical computing entails writing a VGAM family
function for handling data sets where there are a mixture of 0s and
values coming from a continuous positive-valued distribution such as
the lognormal.

Once the function has been written, some data analysis involving
fisheries data should be conducted.  In particular, the catch
data is to be modelled in relation to other variables (fishing effort,
vessel parameters etc) to provide an abundance index.  The software
will be used to analyze catch data collected from the trevally fishery
from the East Northland and Hauraki Gulf areas to generate an annual
index.

Ideally the student can fit regression models competently, knows S4 R
programming and has a sound background in statistical theory.