Applied ARIMA(x) Analysis Model in R

[@yohani1573]

Great video - you make some of the best videos on ML for R that I've seen. Would be very keen to see the SARIMA video!

[@SpencerPaoHere]

Thanks for your support! I was thinking about doing SARIMA models in the future. I'll put that on the list :)

[@yohani1573]

@@SpencerPaoHere Awesome! Quick question - what do you think are the best models for time series forecasting with a leading indicator?

[@SpencerPaoHere]

@@yohani1573 It honestly really depends. You would typically have to experiment with a variety of models and evaluate your metrics. If you receive consistent metrics across all of your models, then the simplest model would be the best.
Some tidbits:
If you don't have a lot of observations, then ARIMA / SARIMA or some version of Fourier transforms might suffice.
If you want to go through the brute force approach where you have lots of observations, then maybe some version of RNN/LSTM/GRU might do you some good.

[@yohani1573]

@@SpencerPaoHere Super helpful - thanks! Would love to see some RNN and BSTS videos as well :)

[@SpencerPaoHere]

@@yohani1573 RNN video: ruclips.net/video/FBlPZJrJt9g/видео.html
Interesting ! BSTS is pretty nifty. I'll add that to my list.

[@Jesus_is_Lord_Ever]

Hi thx for the video, may i know where can i get the data set and code you used?

[@SpencerPaoHere]

Hi! No problem :)
I obtained adjusted closing data from Yahoo Finance and the Treasury rates at treasury.gov. I manually downloaded the separate datasets and joined them based on daily dates.
You can check out my Financial scraping video for additional details if you'd want to skip the manual process of downloading data sets :)
The code & data can be seen here: github.com/SpencerPao/ARIMA

[@Jesus_is_Lord_Ever]

@@SpencerPaoHere Thank you very much.