Which model would you use to predict traffic for a restaurant?

Which model would you use to predict how many of which foods would be ordered in a restaurant?

I was thinking about this question a lot, and cannot figure out an answer.

I would need a hourly breakdown of it, and I was thinking a lstm, where the xs being the time, date, day of week, and amount ordered, and I would need the expected orders for the next hour. This model for each food.

The problem with this is …

datascience traffic