Ok you are forecasting the future without any data in the future - a classical timeseries prediction. the test segments has minutes 1 to 15 available as input and the task is to forecast minutes 17, 18, 19, 20, and 21 — the next 5 minutes, with a one-minute embargo (meaning minute 16 is excluded).
When you sort by the time_segment_id column, you’ll notice that each test input segment contains 15 timesteps. The task is to forecast the following 5 minutes, for which no data is available.
'/%3e%3c/svg%3e)
Ok you are forecasting the future without any data in the future - a classical timeseries prediction. the test segments has minutes 1 to 15 available as input and the task is to forecast minutes 17, 18, 19, 20, and 21 — the next 5 minutes, with a one-minute embargo (meaning minute 16 is excluded).
You can check out this starter notebook (rephrases the problem as a prediction problem), copy it, run it and generate a baseline..https://www.kaggle.com/code/juliusmwangi/what-causes-traffic-congestion-in-barbadoes. Then you can reowork to improve with time
Are you still stuck @Alanin?
Yh my test set does bot have minutes 1 to 15 like what you just posted
When you sort by the time_segment_id column, you’ll notice that each test input segment contains 15 timesteps. The task is to forecast the following 5 minutes, for which no data is available.