As I am going to be using TIM (Tangent Information Modeller) to do the forecasting it seemed a good place to start so I have tried to digest as much information as I could. It is very different to using something like sci-kit learn or TensorFlow as there are no user forums or communities that use it – just the official documentation, which is mainly dedicated to the theory and Tim Studio which will only be used for testing. This is going to be a challenge as the configuration settings are not well explained. It is possibly going to take a lot of trial and error to produce anything accurately, and my feeling is at the moment that this would be better done with sci-kit learn of Facebook Prophet, but cannot be as it is essentially a showcase for TIM.
Having played with TIM Studio and sample data there is a fair amount of research that I need to do – the language is very different from what we have learnt on the course – e.g. Backtesting which I think is the same as using a training set and test set. I am unsure at this point whether that is peculiar to TIM or more pertinent to time series analysis.
I have also installed the TIM Python client and run the sample scripts. TIM cannot be installed using PIP so I have set up a virtual environment and done a manual install within that. Running the sample scripts through Jupyter notebooks has given me some ideas as to how it could be used but the problem is in all their samples they are starting with clean data and I suspect that the biggest problem will be ingesting the data from multiple sources and cleaning it.