Smart Air lent this one to me as it’s new on the market. If the manufacturers are willing to lend me the devices I would be happy to test them in the same conditions.
Awesome convenience functions! One thing I would recommend is to add a docstring to each to remind yourself (and others) what the function does and the arguments to pass through. Thanks for sharing!
Good idea! I will work on checking both the time series and ML model forecasts against the real data up until today.
Recently after I posted this article the IMF has been advising MongolBank on how to keep the exchange rate steady, so I will be interested to see how this impacts the performance of the model.
How do you feel about using this method as opposed to pickling a model for use with Flask? I can see the obvious advantage of keeping all code in one place, but can you keep a notebook running for days or weeks?
Very interesting read!
I had no idea you published the data set! I recently published a 50k row crime dataset scraped from crimemap.police.gov.mn.
I will definitely be doing an analysis using this dataset!
Quick question, do you have access to other fields such as the comments or response by the agency/ministry it’s…
Yes my final model that is detailed in part 3 of the series used weather forecasts to predict PM2.5 pollution. It’s a bit tough though as the forecasts can change quite often.