ETC1010: Data Modelling and Computing

Topic 1 covers spreadsheet modelling. Is VERY PRACTICAL, and is almost an excel tutorial. The whole topic is based on making decisions to optimise a certain output variable (eg profit). Lectures involved Lee effectively running through the construction of a model, and its interpretation. You learn a few handy excel functions, concept diagrams and sensitivity analysis. There is also a tiny bit of conditional probability chucked on the end, which isn't very demanding especially if you've done Maths Methods. That being said, the lectures aren't too important in terms of the exam. More on that later

I really enjoyed the lectures. Lee was very engaging, and you could immediately see that the stuff you were being taught was very applicable to the real world. I've already used some of the stuff we used at work.

The first assignment is from this topic, and you have to create a spreadsheet model and interpret it by answering a few questions. It's not too difficult (most people, myself included, tried to make it more complex than it really needed to be) but there are a few things that you can lose easy marks on. This included not identifying/highlighting your decision and output variables in the concept diagram and model; and not referring to specific figures from you model in the report. As in don't just say "Dodo power and gas should be chosen", say "the optimal decision in order to minimise average complaints is to choose Dodo Power and gas, because *insert values from the model here*". If you do these things you should be able to get a fairly good mark.

Tutes are very similar to the assignment, in the sense that you create a model and interpret it. Can be pretty fun if you have a good tute and tutor (Behrooz!!!) and your attendance is marked, so best to show up.

Multiple regression. Lectures were not particularly helpful here to be perfectly honest. Most of the stuff was covered in ETC1000. New parts were prediction intervals, taking the log of the dependent and independent variables, and interaction terms (you don't even have to know how to interpret them, just know that they exist and can change interpretations) all of which is covered thoroughly in the notes. But yeah, all of the hypothesis testing, dummy variables, confidence intervals and coefficient interpretation was covered in ETC1000.

Tutes are no longer marked, and are also no longer helpful as well, unless you have specific questions to ask the tutor. To be fair, I did this unit at the same time as ETC2410 (intro econometrics) so had done all of the new material 3 or 4 weeks prior, so don't really know how helpful the tutes and lectures were for learning this stuff personally. From speaking with other people though they too said they were not particularly beneficial.

All that you need to know for the assignment for this topic is covered in the lecture notes. If you have them next to you while doing it you shouldn't have too many hiccups. Just phrase your interpretations and set out your hypothesis tests the way they do in the notes and you're set.

A very basic introduction to actuarial studies. Seems almost like a sales pitch trying to get more actuary majors. Covers compound interest and time value of money (If you've done AFC1000 you know this stuff already, even further maths and you should be sweet), demography (life tables, population pyramids) and also what insurance is, various types of policies, and introduces a few probability distributions that actuaries may use. Not a great deal of maths, all the formulas are given to you and you basically only have to know the names of the various distributions. A little bit on the principle of equivalence too.

All the topics are very introductory, and can be easily learned from the lecture slides and the past papers. Some people had stopped coming to the lectures after topic 2, and they follow the lecture slides extremely closely.

No assignments for this topic either. Tutes can be worth it, life tables may take a little getting used to so they can help with that. Decide for yourself. I personally didn't get much out of them (still went because my tute was good fun) but others found them to be pretty helpful.

Two 1-hour lectures and one 1.5-hour tutorial/computer lab = 3.5 hours

2013 Semester 2

87