Python Forum

Full Version: Python prediction on historical data and/or external factors.
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Hi all,

I am posting for the first time to ask for guidance and i believe this forum section to be the most appropriate for my request.

I have been recently tasked to prepare a little exercise to assess your technical expertise on data modeling and problem solving creativity in a Bi solution where python and R scripts can be used.

The exercise is about office Planning showing which employee has been in office or working from home on which days over the last couple of weeks. Of course this is fake data only. Based on excel data link I was asked first to provide a BI solution for attendance visualization which I did as i have some knowledge on this specific Bi tool.


I was afterwards tasked to provide if possible to add a Python (or equivalent) supported prediction for next two weeks based on historical data and/or external factors which i can deliver because of time constraints and also because i am relatively new and trying to start learning Python.

The part where I need assistance or guidance is the following:
So because i am not technically able for now to deliver the predictive model ive been asked instead to get back with a concept/idea on how predictive analytics could support. Which technical solution would you propose? What could be interesting questions to be answered? Which external factors could be used?


Can anyone guide tell me if that is even possible with python and if more variable factors are needed for the prediction ? how much effort would it take to build ? and if you can please guide me for documentation i can refer you to provide a sound based alternative solution as described above ?

I hope my posting format is correct and also did not violate any rules.

Thank you in advance.
The simple answer : yes python can do it.
The difficult answer: you need some kind of model. I would start by
looking into the examples from things with names like "machine learning" or "big data".
They are in the business of examining large amounts of historical data and make projections from there.
Find a model/example that looks like yours, and be prepared for doing overtime. :-)
my 2 cts,
Paul