Ok, so I’ve a small bit more clarity on the jobs front. I’ve a list of places that Mark gave me to try out. There’s no real phD here I would think.
Next steps is to get resume in shape over the weekend. That and probability.
Career
I had an idea of pushing myself as a ‘worker of the future’ or a ‘worker for the coming shift in direction’ or something like that, just as a guiding principle for my cover letter.
Research
The representation of the LCA is important as it defines what you can and cannot say about the output. If your LCA is representated as a set of matrices or as a system of linear equations different numerical analysis methods are open to you. Arguably the most flexible representation is that of a computer program. This allows the LCA designer to choose a different underlying representation based on what they would like to say about the output.
All the issues and benefits are downstream of the chosen representation.
The argument as to why do it in a programming language is that its the most flexible representation of the LCA.
Groen 2014
So it seems that they choose a matirx representation and do matrix calculations.
The presentation in Table 3 is similar to what we would be hoping for. In that, assumptions are right beside values. In a matrix representation its like a dynamic, interactive matrix. This is the simple division of product and elementary flows too which is nice and may be a better model to follow.
The statistical language, I’m not too familiar with, things like Standard error. Convergence I’ve only touched on.
Quasi MC and Latin hypercube seem to be more ‘accurate’ than MC with these statistical metrics.