Which subject is more advanced and useful [closed]

Question

I know this is an open-ended question but I'd like to leverage wisdom of the crowd to make an optimal decision.

The case is, I never learned statistics until 2 years ago. Since then, I'm quite attracted to it and want to solidify my knowledge on the field by looking for some courses to enroll in. I came across these three which sound interesting. The problem is I can only afford to do one of them.

So, if you have to choose, which one would you go for:

1. This course uses Statistics: The Art and Science of Learning from Data, 4th Edition 2017 Alan Agresti, Christine Franklin, Bernhard Klingenberg as texbook. Upon successful completion of this subject students should be able to:
   • Identify and justify statistical concepts and tools to analyse a real-world research problem.
   • Use appropriate statistical techniques to conduct exploratory data analysis and present numerical and graphical summaries.
   • Apply inferences from sample data to populations.
   • Explain and relate the assumptions underlying the use of particular statistical techniques and check whether they are appropriate for a given data sample.
   • Conduct statistical analysis and interpret associated computer output.
2. This course uses K.Borovkov, Elements of stochastic modelling. and E. Platen, D. Heath, A benchmark approach to quantitative finance as textbooks. Upon successful completion of this subject students should be able to:
   • Define and illustrate the terms used in probability and stochastic processes;
   • Discuss and demonstrate the techniques of proof used in probability and some of the mathematical derivations that are important in the theory of stochastic processes;
   • State and apply the basic limit theorems of probability;
   • Demonstrate an ability to use mathematical techniques to analyse the behaviour of various stochastic processes, especially the long-run or steady state behaviour;
   • Formulate and solve applied and theoretical problems involving probability and stochastic processes.
3. This course is called Advanced Bayesian Methods. The subject outline doesn't spell out specific outcomes and no textbook nominated.

Ideally, I want to learn how to learn something practical and useful for my role as safety analyst. I want to be able to use machine learning to model something such as how many accidents will happen given the historical data. Your advice would be much appreciated.

Thanks in advance.

Answer 1

You ask "which is more advanced and useful" but we must presume you mean useful to you, not to us. You also say that you can only afford one and that you want to be able to do something "useful". Unfortunately, if we combine those goals with your comments about how much statistics you know (i.e. not much) then they conflict.

One course is not enough.

If I were you, I would take course 1. First, while I don't know that particular text, Agresti is a very good writer. He is a very clear writer, given the level of whatever book or article he is writing. Second, knowing the basics really well is very important (and lacking in many people who call themselves data scientists).

But perhaps I am misreading your statements about how much statistics you know. You mention "hearing about" statistics 2 years ago. But what were you doing for those 2 years? If you were taking statistics courses, then surely one of them covered the sort of material in course 1. But, if that's the case, you would not be asking this question.

Course 2 is going to be much more mathematical. I'd bet that there are mathematical prerequisites for that course. Do you have them? Were/are you a math major? In any case, I think this sort of theory course is less useful for practical work (other people may disagree).

Course 3 is specialized. Bayesian methods are very popular and useful, but taking course 3 without course 1 (or its equivalent) is like trying to build an apartment building starting on the 10th floor.