By Charles I N I•
The instructor lethargically fits 15 mins worth of content in a 2 hrs video where he copies paste codes from another notebook which is also full of typos and he has to fix it live (and it's not quick by any means!!).
The content is not generalizable, the timeseries chosen was already stationary and did not require any differencing which is not as common. Also, the LSTM forecst model was chosen to be a single-step and not multi-step. Which is easy and can be found everywhere online.
The instructor also says in the intriduction that you need absolutly no background to do this. the he goes on and uses Autolag Adfuller test, ACF and PACF plots without explaining what they mean and how they should be interpreted. He also uses batch size, window size and other terms related to LSTM and does not explain what they are.
Overall, the content was poor and the presentation was even worse.
By Salvatore T•
I think this course goes way too fast. In a Guided Project one is supposed to code along with the instructor, but this is definitely not the case.