How is it forecasting if future features are unknown, Isnt it just regression?

[f-izzat]

• Today is 2022-08-30, I have the high value for this.
• I have followed the Forecasting stocks notebook
• Say I would like to forecast the value for 2022-08-31, but I dont have the value for this in order to extract my X features
• Why would I want a model to predict the high value for today when I already have it?
• Title for forecasting should be changed to 'regression' unless there is a way to predict future values and not just today

[nils-braun]

I assume you are referring to this notebook. The procedure described there is used to predict the value on 2022-08-31 using data up to and including 2022-08-30. The notebook is using shift operation to accomplish this.
Of course, it is still a regression (so you are right), because the target is real-valued (and not a classification problem). It is however also a prediction because I am using values from the past to predict values of the future. You could change the shift parameter from -1 to -7 to try predicting the values 7 days into the future.