Preliminary Action Plan

[msaltieri]

Guys,
thank you for your support and collaboration.

As already agreed, I think the activities and milestones that have to be reached in this preliminary step are the following:

• build a robust framework for basic objects: market data, indicators, deals, portfolios, backtesting (@msaltieri);
• brainstorm different high level options for theoretical model in term of response variable definition and decision making (@msaltieri @mattiapenati @crimarchesini);
• design the theoretical statistical model for the response variable, seen as the probability for the signal to exceed a specific number of standard deviations in a certain time interval (@mattiapenati and @msaltieri);
• experiment very basic strategies playing around with the R interface and commands (@crimarchesini);
• test more sophisticated strategies based on machine learning (supervised or unsupervised) applied to filtered signals (i.e. Fourier transform) (@mattiapenati).

If you think I missed something, please feel free to add a comment below.

[crimarchesini]

Great - I would only add

• brainstorm different high level options for theoretical model in term of response variable definition and decision making (alternatives to point 2 above) @msaltieri @mattiapenati @crimarchesini.

[msaltieri]

Great! I've just updated the original comment so that we can handle in the same place all the actions to take.

I've also added in the home page (Collaboration section) some useful links to cheatsheets and tips & trick pages on the R programming language. Hope it helps!

I'm actually working on a structure which will be able to define the ledger object, i.e. the deals' container. More details will follow shortly.