In this quick power bi tutorial video I will show you how to add forecasting to your timeseries based line graphs. This video is a complete walk-through and I will walk you through every step from selecting the number of forecast points or periods to adding in seasonality correctly. For this to work correctly you need to have time-based columnar data. If you don't the forecast function will not show.
Get a free stock like Apple, Ford, or Facebook worth up to $500 just by signing up and joining Robinhood using my link: join.robinhood.com/davidm16707
To best show it I have two visuals one is with the forecasting and trend lines and is more of a finished visualization. The other one is a base line graph and I show you how to take that and make it look exactly like the other one. I cover everything from proper field selection to adding the forecast, 95% confidence intervals, seasonality, color selection, and basically everything you need to make your power bi visuals pop.
As stated in the video, do not use this for limited data with few data points. As you can see in the video, I am using the commercial tool sales dataset from Kaggle and this has 1.5 years of weekly data, so there is plenty of data here with which to make some reliable forecasts. Now, remember, with seasonality for exponential smoothing (which is the model used by Power BI) it is recommended that you use about 1/4th of the known data points as your seasonality value. You can also leave it at auto and may find that score and auto to be the same.
This video will get you making some really cool forecasted line graphs that you can add to your next Power Bi dashboard, report or analysis and make them standout. You can also take these graphs and filter the data backwards and plot the forecast on previous data then compare it to the data and graph with the filtered data added back in and see and measure how accurate your forecasts were. I will show you how to do this in an upcoming video.
I hope you found this power bi tutorial video on how to make amazing forecasts to be helpful and interesting.
Please take a moment to subscribe, like and share. Also, don't forget to click the bell so you get notified every time I publish a new great video on data analytics and data science stuff.
Thanks again and God bless!
16 июл 2019
